@blockrun/franklin 3.42.2 → 3.43.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +5 -5
- package/dist/agent/commands.js +2 -2
- package/dist/agent/compact.js +6 -3
- package/dist/agent/context.js +1 -1
- package/dist/agent/evaluator.js +2 -1
- package/dist/agent/llm.js +4 -3
- package/dist/agent/loop.js +82 -14
- package/dist/agent/nemotron-prose-stripper.d.ts +26 -0
- package/dist/agent/nemotron-prose-stripper.js +39 -1
- package/dist/agent/tokens.js +22 -5
- package/dist/agent/turn-analyzer.js +18 -6
- package/dist/agent/verification.js +2 -1
- package/dist/brain/extract.js +2 -1
- package/dist/commands/slack.js +2 -1
- package/dist/commands/social.js +2 -1
- package/dist/commands/telegram.js +2 -1
- package/dist/free-models.constants.d.ts +28 -0
- package/dist/free-models.constants.js +33 -0
- package/dist/free-models.d.ts +320 -0
- package/dist/free-models.js +407 -0
- package/dist/gateway-models.js +15 -3
- package/dist/learnings/extractor.js +3 -2
- package/dist/plugin-sdk/workflow.js +5 -2
- package/dist/pricing.js +18 -9
- package/dist/proxy/fallback.d.ts +7 -0
- package/dist/proxy/fallback.js +14 -1
- package/dist/proxy/server.js +15 -2
- package/dist/router/index.d.ts +3 -1
- package/dist/router/index.js +140 -89
- package/dist/router/vision.js +35 -10
- package/dist/serve/server.js +1 -4
- package/dist/tools/moa.js +9 -3
- package/dist/tools/subagent.js +6 -5
- package/dist/ui/model-picker.js +75 -74
- package/package.json +2 -2
package/README.md
CHANGED
|
@@ -286,7 +286,7 @@ Every tool call is itemized. Every token is priced. When the wallet hits zero, F
|
|
|
286
286
|
|
|
287
287
|
## Smart Router
|
|
288
288
|
|
|
289
|
-
**<!-- br:models.chatVisible -->
|
|
289
|
+
**<!-- br:models.chatVisible -->76<!-- /br:models.chatVisible --> models. One decision. Zero guesswork.**
|
|
290
290
|
|
|
291
291
|
You don't pick models. Franklin picks for you.
|
|
292
292
|
|
|
@@ -345,7 +345,7 @@ You don't subscribe to electricity, you pay for what you use. Franklin brings th
|
|
|
345
345
|
|
|
346
346
|
### 🧠 Multi-model is the future
|
|
347
347
|
|
|
348
|
-
No single model is best at everything. Sonnet writes better code, Gemini handles longer context, DeepSeek costs 20x less for simple tasks. The Smart Router routes every request to the optimal model in <1ms — <!-- br:savings.autoVsBaselinePct -->
|
|
348
|
+
No single model is best at everything. Sonnet writes better code, Gemini handles longer context, DeepSeek costs 20x less for simple tasks. The Smart Router routes every request to the optimal model in <1ms — <!-- br:savings.autoVsBaselinePct -->84<!-- /br:savings.autoVsBaselinePct -->% cheaper than pinning Claude Opus 5 for every request.
|
|
349
349
|
|
|
350
350
|
</td>
|
|
351
351
|
<td width="33%" valign="top">
|
|
@@ -367,7 +367,7 @@ No email. No phone. No KYC. Your Base or Solana address is your account — port
|
|
|
367
367
|
| Writes code | ✅ | ✅ | ⚠️ | ✅ |
|
|
368
368
|
| **Spends money for you** | ❌ | ❌ | ❌ | ✅ **USDC wallet, x402** |
|
|
369
369
|
| **Buys data + APIs + images + search** | ❌ | ❌ | ❌ | ✅ **55+ APIs, one wallet** |
|
|
370
|
-
| Picks best model per task | ❌ single-vendor | ❌ plan-tied | ❌ | ✅ **Smart Router, <!-- br:models.chatVisible -->
|
|
370
|
+
| Picks best model per task | ❌ single-vendor | ❌ plan-tied | ❌ | ✅ **Smart Router, <!-- br:models.chatVisible -->76<!-- /br:models.chatVisible --> models** |
|
|
371
371
|
| Pricing model | Subscription | Subscription | Subscription | **YOPO** — per outcome, USDC |
|
|
372
372
|
| Monthly fee | $20–$200 | $20–$40 | $20+ | **$0** |
|
|
373
373
|
| Rate-limited | Yes | Yes | Yes | No — limited only by wallet |
|
|
@@ -395,14 +395,14 @@ Ask "what's BTC looking like?" — Franklin fetches live price data, computes RS
|
|
|
395
395
|
**🎨 AI image generation**
|
|
396
396
|
Ask "generate a logo" — Franklin calls DALL-E / GPT Image, saves the result locally, paid from your wallet.
|
|
397
397
|
|
|
398
|
-
**🧠 <!-- br:models.chatVisible -->
|
|
398
|
+
**🧠 <!-- br:models.chatVisible -->76<!-- /br:models.chatVisible --> models via one wallet**
|
|
399
399
|
Anthropic, OpenAI, Google, xAI, DeepSeek, GLM, Kimi, Minimax, NVIDIA free tier. One wallet, one interface, automatic fallback.
|
|
400
400
|
|
|
401
401
|
**💳 x402 micropayments (YOPO)**
|
|
402
402
|
HTTP 402 native. Every paid action is a signed USDC micropayment via EIP-712 — non-custodial, your keys never leave your machine. YOPO: you pay only for outcomes.
|
|
403
403
|
|
|
404
404
|
**🧠 Learned model router**
|
|
405
|
-
Trained on 2M+ real requests. Classifies your task and picks the best model from <!-- br:models.chatVisible -->
|
|
405
|
+
Trained on 2M+ real requests. Classifies your task and picks the best model from <!-- br:models.chatVisible -->76<!-- /br:models.chatVisible --> LLMs. Four profiles (auto/eco/premium/free). Adapts to your usage over time.
|
|
406
406
|
|
|
407
407
|
</td>
|
|
408
408
|
<td width="50%" valign="top">
|
package/dist/agent/commands.js
CHANGED
|
@@ -14,6 +14,7 @@ import { estimateHistoryTokens, getAnchoredTokenCount, getContextWindow, resetTo
|
|
|
14
14
|
import { forceCompact } from './compact.js';
|
|
15
15
|
import { getStatsSummary, recordUsage } from '../stats/tracker.js';
|
|
16
16
|
import { resolveModel } from '../ui/model-picker.js';
|
|
17
|
+
import { isFreeModelId } from '../free-models.js';
|
|
17
18
|
import { listSessions, loadSessionHistory, } from '../session/storage.js';
|
|
18
19
|
import { matchSkill } from '../skills/invoke.js';
|
|
19
20
|
// ─── Git helpers ──────────────────────────────────────────────────────────
|
|
@@ -1049,8 +1050,7 @@ export async function handleSlashCommand(input, ctx) {
|
|
|
1049
1050
|
ctx.config.baseModel = newModel; // Update recovery target so loop doesn't reset
|
|
1050
1051
|
ctx.config.onModelChange?.(newModel, 'user');
|
|
1051
1052
|
// Warn when switching from free to paid so users know charges start now
|
|
1052
|
-
const
|
|
1053
|
-
const paidWarning = !isFree(newModel)
|
|
1053
|
+
const paidWarning = !isFreeModelId(newModel)
|
|
1054
1054
|
? ` ⚠️ (paid — charges from your wallet per call)`
|
|
1055
1055
|
: '';
|
|
1056
1056
|
ctx.onEvent({ kind: 'text_delta', text: `Model → **${newModel}**${paidWarning}\n` });
|
package/dist/agent/compact.js
CHANGED
|
@@ -5,6 +5,7 @@
|
|
|
5
5
|
*/
|
|
6
6
|
import { existsSync, readFileSync } from 'node:fs';
|
|
7
7
|
import { estimateHistoryTokens, getCompactionThreshold, COMPACTION_SUMMARY_RESERVE, } from './tokens.js';
|
|
8
|
+
import { FREE_DEFAULT_MODEL, isFreeModelId } from '../free-models.js';
|
|
8
9
|
/** Max files to restore after compaction */
|
|
9
10
|
const POST_COMPACT_MAX_FILES = 5;
|
|
10
11
|
/** Max tokens to spend on post-compact file restoration */
|
|
@@ -454,9 +455,11 @@ function formatCompactSummary(raw) {
|
|
|
454
455
|
* so users don't get silent charges when their context fills up.
|
|
455
456
|
*/
|
|
456
457
|
function pickCompactionModel(primaryModel) {
|
|
457
|
-
// Free parent → free compaction (no silent charge)
|
|
458
|
-
|
|
459
|
-
|
|
458
|
+
// Free parent → free compaction (no silent charge). Uses the shared
|
|
459
|
+
// predicate: `startsWith('nvidia/')` would have sent a free-tier user's
|
|
460
|
+
// compaction to the PAID nvidia/kimi-k2.5 without a warning.
|
|
461
|
+
if (isFreeModelId(primaryModel)) {
|
|
462
|
+
return FREE_DEFAULT_MODEL;
|
|
460
463
|
}
|
|
461
464
|
// Use cheapest capable model for summarization to save cost
|
|
462
465
|
// Tier down: opus/pro → sonnet, sonnet → haiku, everything else → flash (cheapest capable)
|
package/dist/agent/context.js
CHANGED
|
@@ -242,7 +242,7 @@ You run on the BlockRun AI Gateway. When the user asks you to "test the BlockRun
|
|
|
242
242
|
- \`GET /.well-known/x402\` — x402 resource list with prices
|
|
243
243
|
|
|
244
244
|
**LLM (POST, x402-paid)**
|
|
245
|
-
- \`POST /v1/chat/completions\` — OpenAI-compatible. Body: \`{ model, messages, stream?, tools?, max_tokens?, temperature? }\`. \`model\` MUST come from \`GET /v1/models\` (real frontier examples on the gateway, verified live 2026-08-
|
|
245
|
+
- \`POST /v1/chat/completions\` — OpenAI-compatible. Body: \`{ model, messages, stream?, tools?, max_tokens?, temperature? }\`. \`model\` MUST come from \`GET /v1/models\` (real frontier examples on the gateway, verified live 2026-08-30: \`anthropic/claude-sonnet-5\`, \`anthropic/claude-opus-5\`, \`openai/gpt-5.6-sol\`, \`deepseek/deepseek-v4-pro\`, \`zai/glm-5.3\`, \`zai/glm-5.3-flash\`, \`xai/grok-4.5\`, \`qwen/qwen3.7-flash\`, \`nvidia/nemotron-3-nano-omni-30b-a3b-reasoning\` (free, the only free id on BOTH the Base and Solana gateways — the other free models are Base-only)). Do NOT invent versions like \`openai/gpt-5.1\` or \`xai/grok-5\` — those don't exist; the gateway 400s with the valid list in the error body, so when in doubt fetch \`GET /v1/models\` first.
|
|
246
246
|
- \`POST /v1/messages\` — Anthropic-compatible. Body: \`{ model, messages, max_tokens, system?, tools? }\`.
|
|
247
247
|
|
|
248
248
|
**Media (POST, x402-paid; GET to poll async jobs)**
|
package/dist/agent/evaluator.js
CHANGED
|
@@ -21,6 +21,7 @@
|
|
|
21
21
|
* PASS) is a v2 concern once we know v1 catches real cases without
|
|
22
22
|
* false-positive noise.
|
|
23
23
|
*/
|
|
24
|
+
import { FREE_DEFAULT_MODEL } from '../free-models.js';
|
|
24
25
|
// ─── Evaluator system prompt ─────────────────────────────────────────────
|
|
25
26
|
//
|
|
26
27
|
// Principle-based, not example-enumerating. Specific tickers or phrasings
|
|
@@ -249,7 +250,7 @@ export function parseGroundingResponse(raw) {
|
|
|
249
250
|
* choice so both quality gates have the same cost profile. Override via
|
|
250
251
|
* `FRANKLIN_EVALUATOR_MODEL` to experiment with accuracy/cost trade-offs. */
|
|
251
252
|
export function evaluatorModel() {
|
|
252
|
-
return process.env.FRANKLIN_EVALUATOR_MODEL ||
|
|
253
|
+
return process.env.FRANKLIN_EVALUATOR_MODEL || FREE_DEFAULT_MODEL;
|
|
253
254
|
}
|
|
254
255
|
// ─── Run grounding check ─────────────────────────────────────────────────
|
|
255
256
|
const MAX_EVAL_TOKENS = 512;
|
package/dist/agent/llm.js
CHANGED
|
@@ -10,6 +10,7 @@ import { routeRequest, parseRoutingProfile } from '../router/index.js';
|
|
|
10
10
|
import { ThinkTagStripper } from './think-tag-stripper.js';
|
|
11
11
|
import { isNemotronProseModel, stripNemotronProse } from './nemotron-prose-stripper.js';
|
|
12
12
|
import { repairAndParseArgs } from './repair/index.js';
|
|
13
|
+
import { FREE_DEFAULT_MODEL } from '../free-models.js';
|
|
13
14
|
// Reasoning-tier models the gateway routes to that reject `tool_choice`
|
|
14
15
|
// outright. Pattern: OpenAI o1/o3 family + DeepSeek's reasoner variant.
|
|
15
16
|
// Add new entries as their 400 errors appear in real sessions; this is
|
|
@@ -465,10 +466,10 @@ export class ModelClient {
|
|
|
465
466
|
// `blockrun/premium` strings (now retired routing profiles) end up at the
|
|
466
467
|
// same place without needing dedicated entries.
|
|
467
468
|
const FALLBACKS = {
|
|
468
|
-
'blockrun/auto':
|
|
469
|
-
'blockrun/free':
|
|
469
|
+
'blockrun/auto': FREE_DEFAULT_MODEL,
|
|
470
|
+
'blockrun/free': FREE_DEFAULT_MODEL,
|
|
470
471
|
};
|
|
471
|
-
return FALLBACKS[model] ||
|
|
472
|
+
return FALLBACKS[model] || FREE_DEFAULT_MODEL;
|
|
472
473
|
}
|
|
473
474
|
/**
|
|
474
475
|
* USDC actually charged for the most recent stream. 0 if no payment
|
package/dist/agent/loop.js
CHANGED
|
@@ -43,7 +43,7 @@ import { writeLiveAgent } from '../session/live-registry.js';
|
|
|
43
43
|
import { estimateCost, OPUS_PRICING } from '../pricing.js';
|
|
44
44
|
import { maybeMidSessionExtract } from '../learnings/extractor.js';
|
|
45
45
|
import { extractMentions, buildEntityContext, loadEntities } from '../brain/store.js';
|
|
46
|
-
import { routeRequest, parseRoutingProfile, getFallbackChain, pickFreeFallback, isVisionModel, messageNeedsVision, pickVisionSibling, markModelUnavailable } from '../router/index.js';
|
|
46
|
+
import { routeRequest, parseRoutingProfile, getFallbackChain, pickFreeFallback, isVisionModel, messageNeedsVision, messagesNeedVision, pickVisionSibling, markModelUnavailable } from '../router/index.js';
|
|
47
47
|
import { recordOutcome } from '../router/local-elo.js';
|
|
48
48
|
import { shouldPlan, getPlanningPrompt, getExecutorModel, isExecutorStuck, toolCallSignature } from './planner.js';
|
|
49
49
|
import { shouldVerify, runVerification } from './verification.js';
|
|
@@ -53,6 +53,7 @@ import { analyzeTurn } from './turn-analyzer.js';
|
|
|
53
53
|
import { evaluateTimeoutRetry } from './retry-policy.js';
|
|
54
54
|
import { MAX_AUTO_CONTINUATIONS_PER_TURN, buildContinuationPrompt, isAutoContinuationDisabled, } from './continuation.js';
|
|
55
55
|
import { createSessionId, appendToSession, updateSessionMeta, pruneOldSessions, loadSessionHistory, loadSessionMeta, } from '../session/storage.js';
|
|
56
|
+
import { freeVisionModel, isFreeModelId } from '../free-models.js';
|
|
56
57
|
/**
|
|
57
58
|
* Atomically replace all elements in a history array.
|
|
58
59
|
* Safer than `history.length = 0; history.push(...)` because if push throws
|
|
@@ -473,9 +474,11 @@ function formatModelSwitch(alias, resolved, reason, newModel) {
|
|
|
473
474
|
*/
|
|
474
475
|
export function isWeakModel(model) {
|
|
475
476
|
const m = model.toLowerCase();
|
|
476
|
-
//
|
|
477
|
-
//
|
|
478
|
-
//
|
|
477
|
+
// Free-tier open models have been observed confabulating tool calls. The
|
|
478
|
+
// whole free pool qualifies, including the non-NVIDIA rungs (poolside,
|
|
479
|
+
// cohere) added when the pool rotated on 2026-08-30.
|
|
480
|
+
if (isFreeModelId(model))
|
|
481
|
+
return true;
|
|
479
482
|
if (m.startsWith('nvidia/'))
|
|
480
483
|
return true;
|
|
481
484
|
if (m.includes('nemotron-ultra'))
|
|
@@ -1392,11 +1395,26 @@ export async function interactiveSession(config, getUserInput, onEvent, onAbortR
|
|
|
1392
1395
|
lastRoutedModel = routing.model;
|
|
1393
1396
|
lastRoutedCategory = routing.category || '';
|
|
1394
1397
|
if (loopCount === 1) {
|
|
1395
|
-
|
|
1398
|
+
// A free-profile vision turn is routed to a TEXT model on purpose:
|
|
1399
|
+
// the gateway has no free model that honours an image (see
|
|
1400
|
+
// router/vision.ts). Say so instead of showing the 👁️ tag over a
|
|
1401
|
+
// model that will answer from the text stub alone — and never fall
|
|
1402
|
+
// back to a paid model, which the free profile must not do.
|
|
1403
|
+
const visionUnavailable = routing.signals.includes('free-vision-unavailable');
|
|
1404
|
+
const visionTag = turnNeedsVision && !visionUnavailable ? ' 👁️' : '';
|
|
1396
1405
|
onEvent({
|
|
1397
1406
|
kind: 'text_delta',
|
|
1398
1407
|
text: `*Auto → ${routing.model}${visionTag}*\n\n`,
|
|
1399
1408
|
});
|
|
1409
|
+
if (visionUnavailable) {
|
|
1410
|
+
onEvent({
|
|
1411
|
+
kind: 'text_delta',
|
|
1412
|
+
text: '*⚠️ No free vision model is available — the free vision ids drop the ' +
|
|
1413
|
+
'image on about half of all calls and answer anyway. This turn answers ' +
|
|
1414
|
+
'from text only, so treat anything said about the image as a guess. ' +
|
|
1415
|
+
'Switch to a paid model (e.g. `/model haiku`) to send it.*\n\n',
|
|
1416
|
+
});
|
|
1417
|
+
}
|
|
1400
1418
|
}
|
|
1401
1419
|
}
|
|
1402
1420
|
else if (turnNeedsVision && !isVisionModel(resolvedModel)) {
|
|
@@ -1409,13 +1427,42 @@ export async function interactiveSession(config, getUserInput, onEvent, onAbortR
|
|
|
1409
1427
|
// overridden). Always emit a visible notice so the user knows their
|
|
1410
1428
|
// pick was overridden and why.
|
|
1411
1429
|
const original = resolvedModel;
|
|
1412
|
-
|
|
1413
|
-
|
|
1414
|
-
|
|
1415
|
-
|
|
1416
|
-
|
|
1417
|
-
|
|
1418
|
-
|
|
1430
|
+
// A free-tier model must NEVER be swapped for a paid one. There is no
|
|
1431
|
+
// free vision model (router/vision.ts), so pickVisionSibling would
|
|
1432
|
+
// return the paid default and start charging a user who explicitly
|
|
1433
|
+
// chose the free tier — silently, mid-turn. Say so and stay free.
|
|
1434
|
+
if (isFreeModelId(original)) {
|
|
1435
|
+
// Stay on the free tier: swap to the chain's free vision model if it
|
|
1436
|
+
// has one, otherwise say so. Never swap a free model for a paid one.
|
|
1437
|
+
const freeVision = freeVisionModel();
|
|
1438
|
+
if (freeVision) {
|
|
1439
|
+
resolvedModel = freeVision;
|
|
1440
|
+
config.model = freeVision;
|
|
1441
|
+
onEvent({
|
|
1442
|
+
kind: 'text_delta',
|
|
1443
|
+
text: `*⚠️ ${original} can't see images — using ${freeVision} for this turn (still free).*\n\n`,
|
|
1444
|
+
});
|
|
1445
|
+
}
|
|
1446
|
+
else {
|
|
1447
|
+
onEvent({
|
|
1448
|
+
kind: 'text_delta',
|
|
1449
|
+
text: `*⚠️ ${original} can't see images, and no free vision model ` +
|
|
1450
|
+
'is reliable right now — the free vision ids drop the image on ' +
|
|
1451
|
+
'about half of all calls and answer anyway. Answering from ' +
|
|
1452
|
+
'text only, so treat anything said about the image as a guess; ' +
|
|
1453
|
+
'switch to a paid model (e.g. `/model haiku`) to send it.*\n\n',
|
|
1454
|
+
});
|
|
1455
|
+
}
|
|
1456
|
+
}
|
|
1457
|
+
else {
|
|
1458
|
+
const visionSwap = pickVisionSibling(original);
|
|
1459
|
+
resolvedModel = visionSwap;
|
|
1460
|
+
config.model = visionSwap;
|
|
1461
|
+
onEvent({
|
|
1462
|
+
kind: 'text_delta',
|
|
1463
|
+
text: `*⚠️ ${original} can't see images — using ${visionSwap} for this turn.*\n\n`,
|
|
1464
|
+
});
|
|
1465
|
+
}
|
|
1419
1466
|
}
|
|
1420
1467
|
// The pre-routing cap was calculated from the virtual profile name.
|
|
1421
1468
|
// Re-apply it to the concrete pick so capability filtering and the
|
|
@@ -1463,6 +1510,22 @@ export async function interactiveSession(config, getUserInput, onEvent, onAbortR
|
|
|
1463
1510
|
// answer (and waste the spend). Opt-in per config. Ordered last so a
|
|
1464
1511
|
// forced turn intentionally rebuilds the prompt from the base (dropping
|
|
1465
1512
|
// the skill hints above); normal turns keep them.
|
|
1513
|
+
// ── Free vision is a ONE-SHOT ──
|
|
1514
|
+
// llama-3.2-11b-vision reads an image reliably as a bare call (30/30 on
|
|
1515
|
+
// a validated fixture) and confabulates in a tool loop: handed the full
|
|
1516
|
+
// tool set it re-called `Read` on the same path until the repeat guard
|
|
1517
|
+
// stopped the turn, in one run out of two. Once the image is actually in
|
|
1518
|
+
// the history there is nothing left to fetch, so take the tools away and
|
|
1519
|
+
// let it answer. Read still runs on the first pass — this only fires
|
|
1520
|
+
// after an image block exists.
|
|
1521
|
+
if (resolvedModel === freeVisionModel() &&
|
|
1522
|
+
callToolDefs.length > 0 &&
|
|
1523
|
+
messagesNeedVision(history)) {
|
|
1524
|
+
callToolDefs = [];
|
|
1525
|
+
callSystemPrompt = systemPrompt +
|
|
1526
|
+
'\n\nThe image is already included above. Answer the question about it ' +
|
|
1527
|
+
'directly. Do not call any tools.';
|
|
1528
|
+
}
|
|
1466
1529
|
const onFinalTurn = config.forceAnswerOnFinalTurn && loopCount === maxTurns;
|
|
1467
1530
|
const toolBudgetSpent = config.maxToolCalls != null && turnToolCalls >= config.maxToolCalls;
|
|
1468
1531
|
if ((onFinalTurn || toolBudgetSpent) && callToolDefs.length > 0) {
|
|
@@ -1588,8 +1651,13 @@ export async function interactiveSession(config, getUserInput, onEvent, onAbortR
|
|
|
1588
1651
|
if (!hasText && !hasTools && !hasThinking) {
|
|
1589
1652
|
// Free-only recovery chain — a free/empty-response session must NEVER
|
|
1590
1653
|
// fall back to a paid model (would silently charge the wallet). Both
|
|
1591
|
-
// entries are $0
|
|
1592
|
-
|
|
1654
|
+
// entries are $0 free-tier ids; the second is deliberately a different
|
|
1655
|
+
// PROVIDER (poolside), not another NVIDIA id.
|
|
1656
|
+
// Built from the live chain, not a literal list: the second rung used
|
|
1657
|
+
// to be a hardcoded `poolside/laguna-xs-2.1`, which is absent from
|
|
1658
|
+
// some gateways and 400s there — so the rescue was guaranteed to fail
|
|
1659
|
+
// on exactly the chains it was meant to rescue.
|
|
1660
|
+
const EMPTY_FALLBACK_MODELS = getFallbackChain('SIMPLE', 'free');
|
|
1593
1661
|
const nextModel = EMPTY_FALLBACK_MODELS.find(m => m !== config.model && !turnFailedModels.has(m));
|
|
1594
1662
|
if (nextModel && recoveryAttempts < 2 && !config.disableModelFallback) {
|
|
1595
1663
|
recoveryAttempts++;
|
|
@@ -16,6 +16,32 @@
|
|
|
16
16
|
* `thinking` so it can be routed to the thinking display channel; the
|
|
17
17
|
* remainder is the user-facing `answer`.
|
|
18
18
|
*/
|
|
19
|
+
/**
|
|
20
|
+
* Which models can put chain-of-thought in `content`?
|
|
21
|
+
*
|
|
22
|
+
* Scope correction (2026-08-31). A previous revision widened this to the
|
|
23
|
+
* whole free tier on the theory that the pool leaks. Re-measured across token
|
|
24
|
+
* budgets, that was wrong: the "leak" is max_tokens TRUNCATION. A Nemotron
|
|
25
|
+
* reasoning model cut off mid-thought emits the partial trace in `content`
|
|
26
|
+
* with finish_reason "length"; given room it returns a clean answer with
|
|
27
|
+
* finish_reason "stop" and the trace in `reasoning_content`.
|
|
28
|
+
*
|
|
29
|
+
* nemotron-3-nano-30b, same prompt, chat/completions:
|
|
30
|
+
* max_tokens 120 -> "length", content = "Okay, the user is asking..."
|
|
31
|
+
* max_tokens 300 -> "stop", content = the actual answer
|
|
32
|
+
*
|
|
33
|
+
* So the fix is to give free calls room (see router/index.ts's classifier
|
|
34
|
+
* budget and turn-analyzer.ts's), not to scrub output. This stays narrow — the
|
|
35
|
+
* Nemotron family, the only models observed producing untagged prose — because
|
|
36
|
+
* a wider net plus an opener match could truncate a legitimate answer that
|
|
37
|
+
* happens to begin "The user asks...".
|
|
38
|
+
*
|
|
39
|
+
* KNOWN LIMITATION: the real discriminator is `finish_reason === "length"`,
|
|
40
|
+
* which this module never sees — the streaming call site in agent/llm.ts
|
|
41
|
+
* decides whether to strip before the finish reason arrives. Until that is
|
|
42
|
+
* plumbed through, the opener list is the proxy for it, which is why the
|
|
43
|
+
* openers must stay specific enough that an untruncated answer cannot match.
|
|
44
|
+
*/
|
|
19
45
|
export declare function isNemotronProseModel(model: string): boolean;
|
|
20
46
|
export declare function stripNemotronProse(text: string): {
|
|
21
47
|
thinking: string;
|
|
@@ -18,6 +18,18 @@
|
|
|
18
18
|
*/
|
|
19
19
|
const REASONING_OPENERS = [
|
|
20
20
|
/^the user (asks|wants|says|requested|is asking|wants me|wrote|just|said)/i,
|
|
21
|
+
// Observed live 2026-08-30 on truncated (finish_reason "length") replies
|
|
22
|
+
// from nemotron-3-nano-30b and nemotron-3.5-lightning.
|
|
23
|
+
//
|
|
24
|
+
// `/^hmm,?\s/` and a bare `/^we need to/` were tried here and REMOVED: they
|
|
25
|
+
// match ordinary answers ("We need to bump the timeout first…"), and because
|
|
26
|
+
// stripNemotronProse discards everything up to the LAST answer-introducer, a
|
|
27
|
+
// false positive silently deletes real content rather than degrading it. The
|
|
28
|
+
// surviving patterns all name the user or the request explicitly, which an
|
|
29
|
+
// answer addressed TO the user does not do.
|
|
30
|
+
/^here'?s\s+(?:a|my|the)\s+thinking process/i,
|
|
31
|
+
/^user (asks|says|wants|requested|wrote)/i,
|
|
32
|
+
/^we need to (classify|answer|determine|decide|figure out|respond)\b/i,
|
|
21
33
|
/^looking at (this|the)/i,
|
|
22
34
|
/^based on (the|this)/i,
|
|
23
35
|
/^according to/i,
|
|
@@ -38,8 +50,34 @@ const ANSWER_INTRODUCERS = [
|
|
|
38
50
|
/(?:^|[\s.])(?:output|response|answer|reply|token)\s*:\s*/gi,
|
|
39
51
|
/\bi(?:'ll| will| shall)\s+(?:output|respond|say|reply|return|emit|write|give|print)\s+(?:the|a|with|out|to|exactly|back|only)?\s*(?:token|word|answer|response|string|text|output|message)?\s*:?\s*/gi,
|
|
40
52
|
];
|
|
53
|
+
/**
|
|
54
|
+
* Which models can put chain-of-thought in `content`?
|
|
55
|
+
*
|
|
56
|
+
* Scope correction (2026-08-31). A previous revision widened this to the
|
|
57
|
+
* whole free tier on the theory that the pool leaks. Re-measured across token
|
|
58
|
+
* budgets, that was wrong: the "leak" is max_tokens TRUNCATION. A Nemotron
|
|
59
|
+
* reasoning model cut off mid-thought emits the partial trace in `content`
|
|
60
|
+
* with finish_reason "length"; given room it returns a clean answer with
|
|
61
|
+
* finish_reason "stop" and the trace in `reasoning_content`.
|
|
62
|
+
*
|
|
63
|
+
* nemotron-3-nano-30b, same prompt, chat/completions:
|
|
64
|
+
* max_tokens 120 -> "length", content = "Okay, the user is asking..."
|
|
65
|
+
* max_tokens 300 -> "stop", content = the actual answer
|
|
66
|
+
*
|
|
67
|
+
* So the fix is to give free calls room (see router/index.ts's classifier
|
|
68
|
+
* budget and turn-analyzer.ts's), not to scrub output. This stays narrow — the
|
|
69
|
+
* Nemotron family, the only models observed producing untagged prose — because
|
|
70
|
+
* a wider net plus an opener match could truncate a legitimate answer that
|
|
71
|
+
* happens to begin "The user asks...".
|
|
72
|
+
*
|
|
73
|
+
* KNOWN LIMITATION: the real discriminator is `finish_reason === "length"`,
|
|
74
|
+
* which this module never sees — the streaming call site in agent/llm.ts
|
|
75
|
+
* decides whether to strip before the finish reason arrives. Until that is
|
|
76
|
+
* plumbed through, the opener list is the proxy for it, which is why the
|
|
77
|
+
* openers must stay specific enough that an untruncated answer cannot match.
|
|
78
|
+
*/
|
|
41
79
|
export function isNemotronProseModel(model) {
|
|
42
|
-
return /^nvidia\/nemotron
|
|
80
|
+
return /^nvidia\/nemotron/i.test(model);
|
|
43
81
|
}
|
|
44
82
|
export function stripNemotronProse(text) {
|
|
45
83
|
if (!text)
|
package/dist/agent/tokens.js
CHANGED
|
@@ -4,6 +4,7 @@
|
|
|
4
4
|
* Anchors to actual API counts when available, estimates on top for new messages.
|
|
5
5
|
*/
|
|
6
6
|
import { peekGatewayModel, warmGatewayModelsCache } from '../gateway-models.js';
|
|
7
|
+
import { FREE_POOL_CONTEXT_FLOOR } from '../free-models.js';
|
|
7
8
|
const DEFAULT_BYTES_PER_TOKEN = 4;
|
|
8
9
|
/**
|
|
9
10
|
* Model-specific bytes-per-token ratios for more accurate estimation.
|
|
@@ -281,17 +282,33 @@ const MODEL_CONTEXT_WINDOWS = {
|
|
|
281
282
|
'moonshot/kimi-k2.5': 128_000,
|
|
282
283
|
'minimax/minimax-m3': 1_000_000,
|
|
283
284
|
'minimax/minimax-m2.7': 128_000,
|
|
284
|
-
//
|
|
285
|
-
//
|
|
286
|
-
//
|
|
285
|
+
// Free tier (refreshed 2026-08-30 — the gateway rotated the whole pool).
|
|
286
|
+
//
|
|
287
|
+
// Every current free id is pinned to FREE_POOL_CONTEXT_FLOOR rather than its
|
|
288
|
+
// catalogued window, deliberately. A request for the 1M-context default is
|
|
289
|
+
// served by the 131K nemotron-3-nano-30b whenever the pool is under load,
|
|
290
|
+
// with no signal to the client — so sizing compaction off 1M would build a
|
|
291
|
+
// prompt the substitute cannot accept, and it would fail mid-session on a
|
|
292
|
+
// user who never picked the substitute. The catalogued windows are recorded
|
|
293
|
+
// in free-models.ts (FREE_MODEL_CONTEXT_WINDOWS) for display.
|
|
294
|
+
//
|
|
295
|
+
// Dead and substituted ids are kept so legacy session records still get a
|
|
296
|
+
// sane window.
|
|
297
|
+
'nvidia/nemotron-3-ultra-550b': FREE_POOL_CONTEXT_FLOOR, // catalogued 1M; pool floor applies
|
|
298
|
+
'nvidia/nemotron-3-nano-30b': FREE_POOL_CONTEXT_FLOOR,
|
|
299
|
+
'nvidia/nemotron-3.5-lightning': FREE_POOL_CONTEXT_FLOOR, // catalogued 1M; pool floor applies
|
|
300
|
+
'nvidia/nemotron-3-nano-omni-30b-a3b-reasoning': FREE_POOL_CONTEXT_FLOOR,
|
|
301
|
+
'nvidia/llama-3.2-11b-vision': 128_000,
|
|
302
|
+
'cohere/north-mini-code': FREE_POOL_CONTEXT_FLOOR,
|
|
303
|
+
'poolside/laguna-xs-2.1': FREE_POOL_CONTEXT_FLOOR,
|
|
287
304
|
'nvidia/qwen3-next-80b-a3b-instruct': 262_144, // NVIDIA EOL 410, 2026-07-27
|
|
288
305
|
'nvidia/qwen3.5-122b-a10b': 131_072,
|
|
289
|
-
'nvidia/nemotron-3-nano-omni-30b-a3b-reasoning': 256_000,
|
|
290
306
|
'nvidia/mistral-nemotron': 131_072,
|
|
291
307
|
'nvidia/step-3.7-flash': 131_072,
|
|
292
308
|
'nvidia/seed-oss-36b': 131_072,
|
|
293
|
-
'nvidia/nemotron-nano-9b-v2': 131_072,
|
|
309
|
+
'nvidia/nemotron-nano-9b-v2': 131_072,
|
|
294
310
|
'nvidia/nemotron-nano-12b-v2-vl': 131_072,
|
|
311
|
+
'nvidia/nemotron-3-super-120b-a12b-free': 131_072,
|
|
295
312
|
'nvidia/llama-4-maverick': 131_072,
|
|
296
313
|
'nvidia/mistral-large-3-675b': 131_072,
|
|
297
314
|
// Qwen (paid) — Max tier is 1M ctx; the generic `qwen` fallback below is
|
|
@@ -26,6 +26,7 @@
|
|
|
26
26
|
* - 30s in-memory cache keyed on the three anchors so back-to-back near-
|
|
27
27
|
* identical turns don't re-pay the latency.
|
|
28
28
|
*/
|
|
29
|
+
import { FREE_DEFAULT_MODEL } from '../free-models.js';
|
|
29
30
|
/**
|
|
30
31
|
* Safe default returned when the analyzer call fails (timeout, parse error,
|
|
31
32
|
* gateway down). Chosen to be neutral:
|
|
@@ -45,8 +46,17 @@ const CONSERVATIVE_DEFAULT = {
|
|
|
45
46
|
const MAX_CURRENT_CHARS = 800;
|
|
46
47
|
const MAX_PREV_REPLY_CHARS = 300;
|
|
47
48
|
const MAX_GOAL_CHARS = 200;
|
|
48
|
-
|
|
49
|
-
|
|
49
|
+
// 2.5s did not leave a reasoning model room to finish and emit its JSON; the
|
|
50
|
+
// abort landed mid-thought and the analyzer silently degraded. See
|
|
51
|
+
// MAX_ANALYZER_TOKENS below.
|
|
52
|
+
const TIMEOUT_MS = 8_000;
|
|
53
|
+
// Raised from 128 on 2026-08-31. The free pool is reasoning models now: cut
|
|
54
|
+
// off mid-thought they emit a partial trace instead of the single-line JSON
|
|
55
|
+
// this parser needs, and a parse failure here is SILENT — it disables pushback
|
|
56
|
+
// handling, ticker/live-data prefetch and planning detection with no error
|
|
57
|
+
// surfaced. Same truncation bug the router classifier had; fixing it there and
|
|
58
|
+
// not here left two thirds of the lesson unapplied. Costs nothing at $0.
|
|
59
|
+
const MAX_ANALYZER_TOKENS = 1_024;
|
|
50
60
|
const CACHE_TTL_MS = 30_000;
|
|
51
61
|
const CACHE_MAX_SIZE = 64;
|
|
52
62
|
// ─── Analyzer prompt ────────────────────────────────────────────────────
|
|
@@ -54,10 +64,12 @@ const CACHE_MAX_SIZE = 64;
|
|
|
54
64
|
// Design: one compact prompt, a few precise examples, instruct the model to
|
|
55
65
|
// emit a single-line JSON. The backbone must produce plain-text structured
|
|
56
66
|
// output under tight max_tokens (thinking-first models leave text empty).
|
|
57
|
-
//
|
|
58
|
-
//
|
|
59
|
-
//
|
|
60
|
-
|
|
67
|
+
// Follows FREE_DEFAULT_MODEL since 2026-08-31 (the ids this comment used to
|
|
68
|
+
// name were all retired in the 2026-08-30 pool rotation). NOTE: the free pool
|
|
69
|
+
// is now reasoning models that need room — see the TRUNCATION note in
|
|
70
|
+
// src/free-models.ts. MAX_ANALYZER_TOKENS/TIMEOUT_MS below have NOT been
|
|
71
|
+
// re-tuned for that and may truncate; see the review finding on this file.
|
|
72
|
+
const ANALYZER_MODEL_DEFAULT = process.env.FRANKLIN_ANALYZER_MODEL || FREE_DEFAULT_MODEL;
|
|
61
73
|
const ANALYZER_SYSTEM = `You analyze ONE user message for Franklin's routing + prefetch harness. Output ONE LINE of compact JSON — no explanation, no markdown, no code fences.
|
|
62
74
|
|
|
63
75
|
## Fields
|
|
@@ -8,6 +8,7 @@
|
|
|
8
8
|
* If FAIL: injects feedback into conversation so the main agent can fix issues.
|
|
9
9
|
* If PASS: work is considered verified.
|
|
10
10
|
*/
|
|
11
|
+
import { FREE_DEFAULT_MODEL } from '../free-models.js';
|
|
11
12
|
// ─── Verification System Prompt ───────────────────────────────────────────
|
|
12
13
|
const VERIFICATION_PROMPT = `You are a VERIFICATION agent. Your job is NOT to confirm that code works — it is to TRY TO BREAK IT.
|
|
13
14
|
|
|
@@ -113,7 +114,7 @@ export async function runVerification(history, handlers, client, config) {
|
|
|
113
114
|
];
|
|
114
115
|
config.onEvent?.({ kind: 'text_delta', text: '\n*Verifying...*\n' });
|
|
115
116
|
// Use agent-tested free model for verification.
|
|
116
|
-
const verificationModel =
|
|
117
|
+
const verificationModel = FREE_DEFAULT_MODEL;
|
|
117
118
|
try {
|
|
118
119
|
// Simple single-turn verification call
|
|
119
120
|
const response = await client.complete({
|
package/dist/brain/extract.js
CHANGED
|
@@ -3,6 +3,7 @@
|
|
|
3
3
|
* Uses cheap model to detect people, projects, companies from conversation.
|
|
4
4
|
*/
|
|
5
5
|
import { loadEntities, saveEntities, upsertEntity, addObservation, upsertRelation, isJunkEntityName, } from './store.js';
|
|
6
|
+
import { FREE_DEFAULT_MODEL } from '../free-models.js';
|
|
6
7
|
// Last entry is the free safety net. nvidia/nemotron-super-49b held it until
|
|
7
8
|
// 2026-08-20, by which point it had left the catalog entirely — the free pool
|
|
8
9
|
// answers for it with a substitute, so the "free fallback" was neither the
|
|
@@ -11,7 +12,7 @@ import { loadEntities, saveEntities, upsertEntity, addObservation, upsertRelatio
|
|
|
11
12
|
const EXTRACTION_MODELS = [
|
|
12
13
|
'google/gemini-2.5-flash-lite',
|
|
13
14
|
'google/gemini-2.5-flash',
|
|
14
|
-
|
|
15
|
+
FREE_DEFAULT_MODEL,
|
|
15
16
|
];
|
|
16
17
|
const VALID_TYPES = new Set(['person', 'project', 'company', 'product', 'concept']);
|
|
17
18
|
const BRAIN_PROMPT = `You are analyzing a conversation between a user and an AI agent. Extract entities (people, projects, companies, products, concepts) mentioned in the conversation.
|
package/dist/commands/slack.js
CHANGED
|
@@ -16,6 +16,7 @@ import { connectMcpServers, disconnectMcpServers } from '../mcp/client.js';
|
|
|
16
16
|
import { loadConfig } from './config.js';
|
|
17
17
|
import { runSlackBot } from '../channel/slack.js';
|
|
18
18
|
import { findLatestSessionByChannel } from '../session/storage.js';
|
|
19
|
+
import { FREE_DEFAULT_MODEL } from '../free-models.js';
|
|
19
20
|
export async function slackCommand(opts) {
|
|
20
21
|
const botToken = process.env.SLACK_BOT_TOKEN;
|
|
21
22
|
const appToken = process.env.SLACK_APP_TOKEN;
|
|
@@ -41,7 +42,7 @@ export async function slackCommand(opts) {
|
|
|
41
42
|
const config = loadConfig();
|
|
42
43
|
const model = opts.model ||
|
|
43
44
|
config['default-model'] ||
|
|
44
|
-
|
|
45
|
+
FREE_DEFAULT_MODEL;
|
|
45
46
|
const workingDir = process.cwd();
|
|
46
47
|
const systemInstructions = assembleInstructions(workingDir, model);
|
|
47
48
|
// Connect MCP servers (Notion, etc.) so the bot exposes their tools — mirrors
|
package/dist/commands/social.js
CHANGED
|
@@ -21,6 +21,7 @@ import { runX } from '../social/x.js';
|
|
|
21
21
|
import { getStats } from '../social/db.js';
|
|
22
22
|
import { loadChain, API_URLS } from '../config.js';
|
|
23
23
|
import { loadConfig as loadAppConfig } from './config.js';
|
|
24
|
+
import { FREE_DEFAULT_MODEL } from '../free-models.js';
|
|
24
25
|
/**
|
|
25
26
|
* Entry point wired from src/index.ts as `franklin social [action] [arg]`.
|
|
26
27
|
*/
|
|
@@ -159,7 +160,7 @@ async function runCommand(options) {
|
|
|
159
160
|
const chain = loadChain();
|
|
160
161
|
const apiUrl = API_URLS[chain];
|
|
161
162
|
const appConfig = loadAppConfig();
|
|
162
|
-
const model = options.model || appConfig['default-model'] ||
|
|
163
|
+
const model = options.model || appConfig['default-model'] || FREE_DEFAULT_MODEL;
|
|
163
164
|
console.log(chalk.dim(` Model: ${model}`));
|
|
164
165
|
console.log('');
|
|
165
166
|
let result;
|
|
@@ -14,6 +14,7 @@ import { allCapabilities } from '../tools/index.js';
|
|
|
14
14
|
import { loadConfig } from './config.js';
|
|
15
15
|
import { runTelegramBot } from '../channel/telegram.js';
|
|
16
16
|
import { findLatestSessionByChannel } from '../session/storage.js';
|
|
17
|
+
import { FREE_DEFAULT_MODEL } from '../free-models.js';
|
|
17
18
|
export async function telegramCommand(opts) {
|
|
18
19
|
const token = process.env.TELEGRAM_BOT_TOKEN;
|
|
19
20
|
const ownerRaw = process.env.TELEGRAM_OWNER_ID;
|
|
@@ -44,7 +45,7 @@ export async function telegramCommand(opts) {
|
|
|
44
45
|
// Model: --model flag > config default > free default.
|
|
45
46
|
const model = opts.model ||
|
|
46
47
|
config['default-model'] ||
|
|
47
|
-
|
|
48
|
+
FREE_DEFAULT_MODEL;
|
|
48
49
|
const workingDir = process.cwd();
|
|
49
50
|
const systemInstructions = assembleInstructions(workingDir, model);
|
|
50
51
|
// Resume the most recent session tagged for THIS owner so a process
|
|
@@ -0,0 +1,28 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Pure free-tier constants — NO imports, no side effects.
|
|
3
|
+
*
|
|
4
|
+
* Split out of free-models.ts so the PUBLIC plugin SDK can name the free
|
|
5
|
+
* default without dragging core internals in behind it. free-models.ts imports
|
|
6
|
+
* gateway-models.ts, which imports config.ts, which does a `readFileSync` and
|
|
7
|
+
* builds `~/.blockrun` paths AT MODULE LOAD. Before this split, importing
|
|
8
|
+
* `@blockrun/franklin/plugin-sdk` just to read DEFAULT_MODEL_TIERS triggered
|
|
9
|
+
* that filesystem I/O — a side effect the SDK never had when the tier was a
|
|
10
|
+
* bare string literal, and a coupling the plugin-sdk header explicitly forbids
|
|
11
|
+
* ("Core stays plugin-agnostic").
|
|
12
|
+
*
|
|
13
|
+
* This keeps ONE definition, which is the point of the whole module: the
|
|
14
|
+
* alternative was copying the literal back into the SDK and re-acquiring the
|
|
15
|
+
* drift this file was created to end.
|
|
16
|
+
*
|
|
17
|
+
* Rationale, measurements and the routing logic all stay in free-models.ts,
|
|
18
|
+
* which re-exports everything here.
|
|
19
|
+
*/
|
|
20
|
+
/**
|
|
21
|
+
* The free model Franklin asks for by default. Chosen for AVAILABILITY, not
|
|
22
|
+
* for being the best — see the full note in free-models.ts.
|
|
23
|
+
*/
|
|
24
|
+
export declare const FREE_DEFAULT_MODEL = "nvidia/nemotron-3-nano-omni-30b-a3b-reasoning";
|
|
25
|
+
/** Free models in preference order, best first. See free-models.ts. */
|
|
26
|
+
export declare const FREE_PREFERENCE_ORDER: readonly string[];
|
|
27
|
+
/** Context budget for ANY free-tier request. See free-models.ts. */
|
|
28
|
+
export declare const FREE_POOL_CONTEXT_FLOOR = 131072;
|
|
@@ -0,0 +1,33 @@
|
|
|
1
|
+
/**
|
|
2
|
+
* Pure free-tier constants — NO imports, no side effects.
|
|
3
|
+
*
|
|
4
|
+
* Split out of free-models.ts so the PUBLIC plugin SDK can name the free
|
|
5
|
+
* default without dragging core internals in behind it. free-models.ts imports
|
|
6
|
+
* gateway-models.ts, which imports config.ts, which does a `readFileSync` and
|
|
7
|
+
* builds `~/.blockrun` paths AT MODULE LOAD. Before this split, importing
|
|
8
|
+
* `@blockrun/franklin/plugin-sdk` just to read DEFAULT_MODEL_TIERS triggered
|
|
9
|
+
* that filesystem I/O — a side effect the SDK never had when the tier was a
|
|
10
|
+
* bare string literal, and a coupling the plugin-sdk header explicitly forbids
|
|
11
|
+
* ("Core stays plugin-agnostic").
|
|
12
|
+
*
|
|
13
|
+
* This keeps ONE definition, which is the point of the whole module: the
|
|
14
|
+
* alternative was copying the literal back into the SDK and re-acquiring the
|
|
15
|
+
* drift this file was created to end.
|
|
16
|
+
*
|
|
17
|
+
* Rationale, measurements and the routing logic all stay in free-models.ts,
|
|
18
|
+
* which re-exports everything here.
|
|
19
|
+
*/
|
|
20
|
+
/**
|
|
21
|
+
* The free model Franklin asks for by default. Chosen for AVAILABILITY, not
|
|
22
|
+
* for being the best — see the full note in free-models.ts.
|
|
23
|
+
*/
|
|
24
|
+
export const FREE_DEFAULT_MODEL = 'nvidia/nemotron-3-nano-omni-30b-a3b-reasoning';
|
|
25
|
+
/** Free models in preference order, best first. See free-models.ts. */
|
|
26
|
+
export const FREE_PREFERENCE_ORDER = [
|
|
27
|
+
'nvidia/nemotron-3-nano-30b', // 5/5 healthy, fastest
|
|
28
|
+
FREE_DEFAULT_MODEL, // the floor this never drops below
|
|
29
|
+
'poolside/laguna-xs-2.1', // different provider
|
|
30
|
+
'cohere/north-mini-code', // different provider; needs max_tokens >= 1200
|
|
31
|
+
];
|
|
32
|
+
/** Context budget for ANY free-tier request. See free-models.ts. */
|
|
33
|
+
export const FREE_POOL_CONTEXT_FLOOR = 131_072;
|