roadmodel 0.2.4__tar.gz → 0.2.6__tar.gz

{roadmodel-0.2.4 → roadmodel-0.2.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: roadmodel
-Version: 0.2.4
+Version: 0.2.6
 Summary: BYO-key CLI that recommends the right AI model, platform, and settings for a prompt.
 Project-URL: Homepage, https://roadmodel.ai
 Project-URL: Repository, https://github.com/nathanramoscfa/roadmodel

{roadmodel-0.2.4 → roadmodel-0.2.6}/docs/catalog.json

@@ -1,9 +1,9 @@
 {
   "schema_version": "2",
-  "generated_at_utc": "2026-05-31T03:03:01Z",
+  "generated_at_utc": "2026-06-13T03:26:33Z",
   "source_doc_sha256": {
-    "model-selector.txt": "923eedff960d815b250a2b7c8489b485d9c8ae44735fcb60d115a44c50a1b65b",
-    "model-tier-cost-scale.md": "22b5d44214a26a2f7be1ae6ccec0d0bcefcd7ed987ace4b385da8db3e5e2c6c9"
+    "model-selector.txt": "3678876fce07458df5b821be0cee45475797adab6c243150ac5630bf7169b23a",
+    "model-tier-cost-scale.md": "7e874453400fa3a6d5d9122f762abe654f9a9f565a307d9391143585b2eba451"
   },
   "models": [
     {
@@ -23,10 +23,52 @@
         "speed": "S"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "AA Intelligence Index 37.1; Output Speed 138.4 tokens/s; AA-Omniscience -4.2; latency leader among Claude family",
+      "headline_benchmarks": "AA Intelligence Index 37.1; Output Speed 143.0 tokens/s; AA-Omniscience -4.2; latency leader among Claude family",
       "pricing_notes": "Hidden by default; Bedrock/Vertex: regional endpoints +10% surcharge; Cache: writes 1.25x, reads 0.1x",
       "best_for": "Speed-optimized lowest-cost Claude model, ideal for simple completions, high-volume repetitive tasks, and latency-sensitive workflows where a lightweight capable response matters more than deep reasoning"
     },
+    {
+      "id": "claude-fable-5",
+      "name": "Fable 5",
+      "input_price_per_1m": 10.0,
+      "output_price_per_1m": 50.0,
+      "cache_read_per_1m": null,
+      "tier_cost": "very-high",
+      "tiers": {
+        "coding": "S",
+        "planning": "S",
+        "agentic": "S",
+        "multimodal": "S",
+        "long-context": "S",
+        "knowledge": "S",
+        "speed": "D"
+      },
+      "jurisdiction": "us",
+      "headline_benchmarks": "AA Intelligence Index 64.9 (#1); HLE 53.3% (#1); Terminal-Bench Hard 62.9 (#1)",
+      "pricing_notes": "Requires data retention approval for Enterprise customers, Teams and individual customers with Privacy Mode enabled; Anthropic stores agent input and output data for harm-prevention processes; this data is not used to train or improve Anthropic models or products; Requests that trip a security guardrail are automatically routed to Claude Opus; About 2x the cost of Claude Opus 4.8; Requires Max Mode on request-based plans",
+      "best_for": "Anthropic's new top-of-line Fable family flagship (no predecessor) \u2014 S-tier across coding, planning, agentic, multimodal, long-context, and knowledge, leading HLE (53.3%) and Terminal-Bench Hard (62.9) with state-of-the-art vision and a 1M default context; about 2x the cost of Opus 4.8 and latency-slow (output ~64 tokens/s), so reserve for the hardest reasoning, agentic, and vision work where maximum capability outweighs cost and speed; security-guardrail trips auto-route to Opus. Tier profile sourced from the catalog cron's 2026-06-11 dry-run reconciliation against the live benchmark sources (\u03c4\u00b2-bench retail not yet published for this model), pending editorial confirmation in the next refresh."
+    },
+    {
+      "id": "codestral",
+      "name": "Codestral",
+      "input_price_per_1m": 0.3,
+      "output_price_per_1m": 0.9,
+      "cache_read_per_1m": null,
+      "tier_cost": "low",
+      "tiers": {
+        "coding": "B",
+        "planning": "C",
+        "agentic": "C",
+        "multimodal": "D",
+        "long-context": "B",
+        "knowledge": "C",
+        "speed": "A"
+      },
+      "jurisdiction": "eu",
+      "headline_benchmarks": "Mistral's code-specialist model; fast low-latency completion / fill-in-the-middle across many languages with a large code context window; specific public benchmark numbers pending",
+      "pricing_notes": "Provider-direct Mistral API per-token pricing (not via the Cursor pool); eu-jurisdiction; prices manually maintained (Mistral publishes no machine-readable price source)",
+      "best_for": "Mistral's dedicated code model \u2014 fast, cheap ($0.90/M output) code completion and fill-in-the-middle under the EU jurisdiction, for autocomplete-style / high-throughput coding loops where an EU operator and low latency matter more than top-tier agentic reasoning. Reached via the `mistral-api` method (provider-direct per-token) with a mistral-api-key; prefer mistral-medium-3.5 for reasoning-heavy coding, codestral for fast bounded completions."
+    },
     {
       "id": "composer-2",
       "name": "Composer 2 (Fast)",
@@ -69,6 +111,48 @@
       "pricing_notes": "-",
       "best_for": "Composer 2's successor at the same output price \u2014 Cursor's purpose-built multi-file agentic editor with frontier-level coding quality and speed-optimized inference; prefer over Composer 2 when both are available since 2.5 supersedes 2 within the same series per the equal-output-price replacement rule (Composer 2 is now Hidden by default on Cursor's pricing page)"
     },
+    {
+      "id": "deepseek-v4-flash",
+      "name": "DeepSeek-V4-Flash",
+      "input_price_per_1m": 0.14,
+      "output_price_per_1m": 0.28,
+      "cache_read_per_1m": null,
+      "tier_cost": "low",
+      "tiers": {
+        "coding": "A",
+        "planning": "B",
+        "agentic": "B",
+        "multimodal": "D",
+        "long-context": "B",
+        "knowledge": "B",
+        "speed": "A"
+      },
+      "jurisdiction": "cn",
+      "headline_benchmarks": "AA Intelligence Index 47 (reasoning, max effort) \u2014 independently measured by Artificial Analysis; SWE-bench Verified 79.0% (DeepSeek-reported); 1M-token context; text-only (no image input); ~90 tokens/s",
+      "pricing_notes": "Provider-direct DeepSeek API per-token pricing (not via the Cursor pool); cache-hit input $0.0028/M",
+      "best_for": "DeepSeek's V4-Flash \u2014 the fast (~90 tokens/s), cheapest DeepSeek variant ($0.28/M output) with a 1M-token context window, for high-throughput / latency-sensitive text and code work under the cn jurisdiction with a deepseek-api-key. Mid-pack general intelligence (Artificial Analysis Intelligence Index 47) paired with a strong, DeepSeek-reported coding result (SWE-bench Verified 79.0%) \u2014 rated coding-A on that basis with the rest of its profile B-tier; text-only (no multimodal). Reached via the `deepseek-api` method (provider-direct per-token). Pick V4-Pro over V4-Flash when reasoning depth or the strongest coding matters; pick V4-Flash when speed and the lowest cost dominate."
+    },
+    {
+      "id": "deepseek-v4-pro",
+      "name": "DeepSeek-V4-Pro",
+      "input_price_per_1m": 0.435,
+      "output_price_per_1m": 0.87,
+      "cache_read_per_1m": null,
+      "tier_cost": "low",
+      "tiers": {
+        "coding": "A",
+        "planning": "A",
+        "agentic": "A",
+        "multimodal": "D",
+        "long-context": "A",
+        "knowledge": "A",
+        "speed": "C"
+      },
+      "jurisdiction": "cn",
+      "headline_benchmarks": "AA Intelligence Index 52 (reasoning, max effort) \u2014 independently measured by Artificial Analysis; SWE-bench Verified 80.6%, LiveCodeBench 93.5, Terminal-Bench 2.0 67.9, Codeforces CodeElo 3206, Putnam-2025 120/120 (DeepSeek-reported); 1M-token context; text-only (no image input); ~46 tokens/s (notably slow)",
+      "pricing_notes": "Provider-direct DeepSeek API per-token pricing (not via the Cursor pool); cache-hit input $0.003625/M",
+      "best_for": "DeepSeek's V4-Pro flagship \u2014 a very low-cost ($0.87/M output), cn-jurisdiction reasoning model with a 1M-token context window and thinking mode on by default. Strong general intelligence (Artificial Analysis Intelligence Index 52, just below Grok 4.3) and a frontier-approaching coding profile (SWE-bench Verified 80.6%, LiveCodeBench 93.5, Codeforces CodeElo 3206) \u2014 these coding numbers are DeepSeek-reported, so it is rated coding-A rather than S pending an independent SWE-bench leaderboard entry. Text-only (no multimodal) and notably slow (~46 tokens/s), so not for latency-sensitive or image work. Reached via the `deepseek-api` method (provider-direct per-token, not the Cursor pool) when the cn jurisdiction is acceptable and a deepseek-api-key is configured \u2014 the cheapest A-tier coding / reasoning option in the catalog."
+    },
     {
       "id": "gemini-2.5-flash",
       "name": "Gemini 2.5 Flash",
@@ -149,7 +233,7 @@
         "speed": "B"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "AA Intelligence Index 57.2 (#3); AA-Omniscience 32.9 (#1); HLE 44.7% (#1); LMArena Text Elo 1482.8 (#5); 1M-token context",
+      "headline_benchmarks": "AA Intelligence Index 57.2 (#3); AA-Omniscience 32.9 (#1); HLE 44.7% (#1); LMArena Text Elo 1482.0 (#4); 1M-token context",
       "pricing_notes": "-",
       "best_for": "True native multimodal understanding (text, image, video, audio, and code in a single pass), 1M-token context optimized for heterogeneous inputs, strong agentic multi-step tool use, and synthesizing insights across large mixed-media datasets or sprawling document corpora \u2014 the obvious choice whenever multimodal or long-context is the primary category"
     },
@@ -170,7 +254,7 @@
         "speed": "S"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "AA Intelligence Index 55.3 (high reasoning); \u03c4\u00b2-bench retail pass_1 45.6 (Gemini 3.5 Flash); Output Speed 227.1 tokens/s",
+      "headline_benchmarks": "AA Intelligence Index 55.3 (high reasoning); \u03c4\u00b2-bench retail pass_1 45.6 (Gemini 3.5 Flash); Output Speed 203.7 tokens/s",
       "pricing_notes": "-",
       "best_for": "Auto-added cheap-tier Google model; pending editorial best-for refinement."
     },
@@ -296,7 +380,7 @@
         "speed": "B"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "AA Intelligence Index 56.8 (#4); LMArena Text Elo 1454.5 (#20); GPT-5.4 (xhigh) Output Speed 80.5 tokens/s; lowest factual error rate among GPT models",
+      "headline_benchmarks": "AA Intelligence Index 56.8 (#4); LMArena Text Elo 1452.5 (#23); GPT-5.4 (xhigh) Output Speed 86.1 tokens/s; lowest factual error rate among GPT models",
       "pricing_notes": "Hidden by default; Requires Max Mode on request-based plans; Agentic and reasoning capabilities; 90% discount on cached input tokens; Fast mode is 15% faster with 2x pricing; Long context (Max Mode) supports up to 1M tokens with 2x input pricing",
       "best_for": "Broadest professional domain expertise (outperforms human specialists in 83% of occupations), native computer-use capability surpassing human baselines, lowest factual error rate among GPT models, and cross-domain knowledge work requiring deep real-world accuracy and grounding"
     },
@@ -317,7 +401,7 @@
         "speed": "A"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "AA Intelligence Index 48.9 (xhigh); Output Speed 160.3 tokens/s; HLE 19.4% (GPT-5-mini)",
+      "headline_benchmarks": "AA Intelligence Index 48.9 (xhigh); Output Speed 180.8 tokens/s; HLE 26.6% (GPT-5.4-mini xhigh)",
       "pricing_notes": "Hidden by default; Smaller, faster variant of GPT-5.4; 90% discount on cached input tokens",
       "best_for": "Lightweight GPT-5.4 variant balancing quality and cost, well-suited for straightforward coding, short-form generation, and high-throughput workloads needing solid GPT reasoning at a fraction of the flagship price"
     },
@@ -359,7 +443,7 @@
         "speed": "D"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "AA Intelligence Index 60.2 (#1); LMArena Text Elo 1464.9 (#16); HLE 44.3%; AA-Omniscience 20.1 (#3)",
+      "headline_benchmarks": "AA Intelligence Index 60.2 (#1); LMArena Text Elo 1462.3 (#16); HLE 44.3%; AA-Omniscience 20.1 (#3)",
       "pricing_notes": "Requires Max Mode on request-based plans; Agentic and reasoning capabilities; More token-efficient than GPT-5.4 on comparable tasks; Improved persistence on long-running tasks; Fast mode is available at higher rates; Long context (Max Mode) supports up to 1M tokens with 2x input pricing",
       "best_for": "OpenAI's most capable frontier model and highest-cost GPT offering, best suited for the most demanding reasoning, long-horizon planning, and tasks where maximum intelligence is required regardless of cost \u2014 strongest single model for hard coding, agentic execution, and reasoning, but verify factual claims due to elevated hallucination"
     },
@@ -381,7 +465,7 @@
       },
       "jurisdiction": "us",
       "headline_benchmarks": "AA Intelligence Index 53.2 (#7); AA-Omniscience 18.3 (#4); HLE 35.0%; LMArena Search Elo 1189.2",
-      "pricing_notes": "Requires Max Mode on request-based plans",
+      "pricing_notes": "Hidden by default; Requires Max Mode on request-based plans",
       "best_for": "Latest Grok release with built-in multi-agent self-verification, configurable reasoning depth, and signature 2M-token context with hallucination-resistant grounding \u2014 leads the low tier on agentic execution and long-context, ideal when massive context, factual accuracy, and aggressive cost efficiency must coexist"
     },
     {
@@ -405,6 +489,69 @@
       "pricing_notes": "Hidden by default",
       "best_for": "Moonshot's affordable mid-volume model \u2014 a non-Google / non-OpenAI / non-Anthropic option at low-tier pricing for cost-conscious code and text generation when provider diversity is desired (vendor-risk hedging, regional preferences); routed via Cursor's pool only \u2014 no direct Moonshot access method is currently enumerated in the access-methods block"
     },
+    {
+      "id": "mistral-large-3",
+      "name": "Mistral Large 3",
+      "input_price_per_1m": 0.5,
+      "output_price_per_1m": 1.5,
+      "cache_read_per_1m": null,
+      "tier_cost": "low",
+      "tiers": {
+        "coding": "C",
+        "planning": "C",
+        "agentic": "C",
+        "multimodal": "D",
+        "long-context": "C",
+        "knowledge": "C",
+        "speed": "B"
+      },
+      "jurisdiction": "eu",
+      "headline_benchmarks": "AA Intelligence Index 23 (independently measured by Artificial Analysis); open-weight Mixture-of-Experts (self-hostable); text-only",
+      "pricing_notes": "Provider-direct Mistral API per-token pricing (not via the Cursor pool); eu-jurisdiction; prices manually maintained (Mistral publishes no machine-readable price source)",
+      "best_for": "Mistral's open-weight Large 3 (MoE) \u2014 an EU-jurisdiction, self-hostable option at very low cost ($1.50/M output) for data-sovereignty workloads or teams that want to run the weights themselves. Artificial Analysis Intelligence Index 23 sits below the frontier and even below Mistral's own Medium 3.5 (Mistral repositioned Large as an open community model) \u2014 pick it for the open-weights / EU-operator profile, not raw capability. Reached via the `mistral-api` method (provider-direct per-token) or self-hosting."
+    },
+    {
+      "id": "mistral-medium-3.5",
+      "name": "Mistral Medium 3.5",
+      "input_price_per_1m": 1.5,
+      "output_price_per_1m": 7.5,
+      "cache_read_per_1m": null,
+      "tier_cost": "low",
+      "tiers": {
+        "coding": "B",
+        "planning": "B",
+        "agentic": "C",
+        "multimodal": "B",
+        "long-context": "B",
+        "knowledge": "B",
+        "speed": "B"
+      },
+      "jurisdiction": "eu",
+      "headline_benchmarks": "AA Intelligence Index 39 (independently measured by Artificial Analysis); unified chat / reasoning / code model with an adjustable reasoning dial (reasoning_effort); multimodal (text + image input)",
+      "pricing_notes": "Provider-direct Mistral API per-token pricing (not via the Cursor pool); eu-jurisdiction; prices manually maintained (Mistral publishes no machine-readable price source)",
+      "best_for": "Mistral's flagship unified model \u2014 the EU-jurisdiction choice for data-sovereignty / EU-regulatory workloads at low cost ($7.50/M output), with adjustable reasoning and multimodal (vision) input. Artificial Analysis Intelligence Index 39 places it mid-pack (below the US/cn frontier such as Gemini 3.1 Pro or DeepSeek V4-Pro) \u2014 pick it when the operator's EU jurisdiction is the deciding constraint, not when raw capability is. Reached via the `mistral-api` method (provider-direct per-token) with a mistral-api-key."
+    },
+    {
+      "id": "mistral-small-4",
+      "name": "Mistral Small 4",
+      "input_price_per_1m": 0.1,
+      "output_price_per_1m": 0.3,
+      "cache_read_per_1m": null,
+      "tier_cost": "low",
+      "tiers": {
+        "coding": "C",
+        "planning": "C",
+        "agentic": "C",
+        "multimodal": "B",
+        "long-context": "C",
+        "knowledge": "C",
+        "speed": "A"
+      },
+      "jurisdiction": "eu",
+      "headline_benchmarks": "AA Intelligence Index 28 (independently measured by Artificial Analysis); compact Mixture-of-Experts unifying the former Small / Magistral / Pixtral / Devstral lines; adjustable reasoning_effort; multimodal (text + image)",
+      "pricing_notes": "Provider-direct Mistral API per-token pricing (not via the Cursor pool); eu-jurisdiction; prices manually maintained (Mistral publishes no machine-readable price source)",
+      "best_for": "Mistral's cheapest fast model ($0.30/M output) \u2014 a small EU-jurisdiction MoE with multimodal input and an optional reasoning dial, for high-throughput / latency-sensitive text and light multimodal work where the EU operator matters and a mistral-api-key is configured. Artificial Analysis Intelligence Index 28 is low, so it is a cost / sovereignty pick rather than a capability pick. Reached via the `mistral-api` method (provider-direct per-token)."
+    },
     {
       "id": "opus-4.7",
       "name": "Opus 4.7",
@@ -422,7 +569,7 @@
         "speed": "D"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "AA Intelligence Index 57.3 (#2); LMArena Text #7 (Elo 1479.1); LMArena WebDev #2 (Elo 1562.4); AA-Omniscience 26.2 (#2)",
+      "headline_benchmarks": "AA Intelligence Index 57.3 (#2); LMArena Text #6 (Elo 1480.4); LMArena WebDev #2 (Elo 1556.5); AA-Omniscience 26.2 (#2)",
       "pricing_notes": "Hidden by default; Requires Max Mode on request-based plans; Up to 1M tokens in Max Mode at the same per-token rates (no long-context surcharge)",
       "best_for": "Deepest abstract and scientific reasoning, highest coherence on long unsupervised multi-step agent chains, best long-context recall at 1M tokens, 128K output ceiling for large single-shot deliverables, and novel problem-solving where high ambiguity demands creative judgment over pattern-matching"
     },
@@ -436,14 +583,14 @@
       "tiers": {
         "coding": "S",
         "planning": "S",
-        "agentic": "A",
+        "agentic": "S",
         "multimodal": "A",
         "long-context": "S",
         "knowledge": "S",
         "speed": "D"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "Auto-added pending editorial tier review; specific benchmark numbers pending next refresh",
+      "headline_benchmarks": "AA Intelligence Index 61.4 (#1); HLE 45.7%; Terminal-Bench Hard 58.3 (top-tier); \u03c4\u00b2-bench retail pass_1 94.4%",
       "pricing_notes": "Requires Max Mode on request-based plans; Fast mode (`claude-opus-4-8-fast`) requires Max Mode; Fast mode is 3x lower per-token pricing than Opus 4.7 fast mode; Up to 1M tokens in Max Mode at the same per-token rates (no long-context surcharge)",
       "best_for": "Anthropic's Opus 4.7 successor at the same very-high tier pricing \u2014 placeholder tier ratings inherited from opus-4.7 pending benchmark coverage; the 3x cheaper fast-mode per-token rate (vs opus-4.7 fast mode) is the headline cost-structure change to surface in the next editorial pass"
     },
@@ -464,7 +611,7 @@
         "speed": "B"
       },
       "jurisdiction": "us",
-      "headline_benchmarks": "AA Intelligence Index 51.7; LMArena WebDev Elo 1522.9 (#7); AA-Omniscience 12.4; top-ranked tool-calling on Anthropic lineage",
+      "headline_benchmarks": "AA Intelligence Index 51.7; LMArena WebDev Elo 1522.2 (#10); AA-Omniscience 12.4; top-ranked tool-calling on Anthropic lineage",
       "pricing_notes": "Requires Max Mode on request-based plans; Up to 1M tokens in Max Mode at the same per-token rates (no long-context surcharge)",
       "best_for": "Top-ranked tool-calling and agentic execution globally, near-Opus coding quality at 2-3x the speed, strong mathematical reasoning (89% MATH), and complex but well-structured tasks needing reliable high-throughput multi-step implementation"
     }
@@ -479,6 +626,7 @@
       "requires": "anthropic-api-key",
       "supports_models": [
         "claude-4.5-haiku",
+        "claude-fable-5",
         "opus-4.7",
         "opus-4.8",
         "sonnet-4.6"
@@ -514,13 +662,14 @@
       "requires": "claude-max-subscription OR anthropic-api-key",
       "supports_models": [
         "claude-4.5-haiku",
+        "claude-fable-5",
         "opus-4.7",
         "opus-4.8",
         "sonnet-4.6"
       ],
       "exposes_max_mode": "no",
       "exposes_thinking": "yes",
-      "best_for": "Default for Claude coding or terminal tasks when a claude.ai Max subscription is active \u2014 $0 marginal cost until the Max budget is exhausted, full tool-use surface, runs as a CLI and as an IDE extension inside Cursor. Heavy Opus usage that would cost over $1,000/mo on per-token API is fully covered by a $100/mo Max plan."
+      "best_for": "Default for Claude coding or terminal tasks when a claude.ai Max subscription is active \u2014 $0 marginal cost until the Max budget is exhausted, full tool-use surface, runs as a CLI and as an IDE extension inside Cursor. Heavy Opus usage that would cost over $1,000/mo on per-token API is fully covered by a $100/mo Max plan. Exposes the full `/effort` dial (low/medium/high/xhigh/max \u2014 Opus 4.6 and Sonnet 4.6 top out at max with no xhigh step; Opus 4.7, Opus 4.8, and Fable 5 expose the full range) plus Ultracode (session-wide xhigh + Dynamic Workflows) and the per-turn `ultrathink` keyword."
     },
     {
       "id": "claude-web",
@@ -569,6 +718,7 @@
       "requires": "cursor-pro-or-ultra-subscription",
       "supports_models": [
         "claude-4.5-haiku",
+        "claude-fable-5",
         "composer-2",
         "composer-2.5",
         "gemini-2.5-flash",
@@ -595,6 +745,21 @@
       "exposes_thinking": "no",
       "best_for": "Cursor IDE \u2014 single Platform covering both UI modes (Composer for multi-file autonomous editing; Chat for interactive model-picker). The operator picks the mode at task time based on the chosen Model: composer-2 / composer-2.5 imply Composer mode; frontier models (opus-4.7, gpt-5.5, sonnet-4.6, etc.) imply Chat mode. Cursor's own Auto and Premium routing modes are deliberately NOT enumerated as roadmodel-recommendable models because their routing is opaque (see `jurisdiction-context` for the rationale) \u2014 operators who want routing behavior pick a specific fixed model and let Cursor's pool handle the call. All routes through the $0-marginal Cursor pool. Defer to claude-code when the chosen model is Claude and claude.ai Max is active (Max budget is cheaper marginal cost than burning Cursor pool tokens on Claude calls that have a dedicated Anthropic subscription path)."
     },
+    {
+      "id": "deepseek-api",
+      "name": "DeepSeek API",
+      "provider": "deepseek",
+      "provider_jurisdiction": "cn",
+      "billing": "per-token",
+      "requires": "deepseek-api-key",
+      "supports_models": [
+        "deepseek-v4-flash",
+        "deepseek-v4-pro"
+      ],
+      "exposes_max_mode": "no",
+      "exposes_thinking": "yes",
+      "best_for": "Direct DeepSeek API access (provider-direct per-token; OpenAI-format at api.deepseek.com and Anthropic-format at api.deepseek.com/anthropic) for the deepseek-v4 models \u2014 cost-conscious coding / reasoning / long-context (1M) work when the cn jurisdiction is acceptable and a deepseek-api-key is configured. Exposes the full thinking dial (toggle + reasoning_effort `high`/`max`). Not routed via the Cursor pool. cn-jurisdiction: excluded by the default allowed-jurisdictions list unless the user opts into cn."
+    },
     {
       "id": "gemini-app",
       "name": "Gemini (web / app)",
@@ -606,7 +771,8 @@
         "gemini-2.5-flash",
         "gemini-3-flash",
         "gemini-3-pro",
-        "gemini-3.1-pro"
+        "gemini-3.1-pro",
+        "gemini-3.5-flash"
       ],
       "exposes_max_mode": "no",
       "exposes_thinking": "yes",
@@ -623,7 +789,8 @@
         "gemini-2.5-flash",
         "gemini-3-flash",
         "gemini-3-pro",
-        "gemini-3.1-pro"
+        "gemini-3.1-pro",
+        "gemini-3.5-flash"
       ],
       "exposes_max_mode": "no",
       "exposes_thinking": "yes",
@@ -640,12 +807,30 @@
         "gemini-2.5-flash",
         "gemini-3-flash",
         "gemini-3-pro",
-        "gemini-3.1-pro"
+        "gemini-3.1-pro",
+        "gemini-3.5-flash"
       ],
       "exposes_max_mode": "no",
       "exposes_thinking": "yes",
       "best_for": "Programmatic / scripted Gemini use with a Google API key. Pay-per-token at Google's published rates. Powers the roadmodel SaaS free-tier surfaces (/recommend on Gemini 2.5 Flash; /roadmap on Gemini 2.5 Flash with 3.1 Pro escalation)."
     },
+    {
+      "id": "mistral-api",
+      "name": "Mistral API",
+      "provider": "mistral",
+      "provider_jurisdiction": "eu",
+      "billing": "per-token",
+      "requires": "mistral-api-key",
+      "supports_models": [
+        "codestral",
+        "mistral-large-3",
+        "mistral-medium-3.5",
+        "mistral-small-4"
+      ],
+      "exposes_max_mode": "no",
+      "exposes_thinking": "yes",
+      "best_for": "Direct Mistral API access (provider-direct per-token; La Plateforme at api.mistral.ai) for the Mistral models \u2014 the EU-jurisdiction option for data-sovereignty / EU-regulatory workloads at low cost. Exposes a reasoning dial on the unified models (Mistral Small 4 / Medium 3.5) via the `reasoning_effort` parameter. Not routed via the Cursor pool. eu-jurisdiction is in the default allowed-jurisdictions list, so Mistral surfaces for any user with a mistral-api-key configured (no jurisdiction opt-in required, unlike cn providers)."
+    },
     {
       "id": "openai-api",
       "name": "OpenAI API",
@@ -697,6 +882,7 @@
       "provider": "Anthropic",
       "tier": "Claude Pro",
       "monthly_usd": 20.0,
+      "annual_usd": 200.0,
       "surface_funded": [
         "claude-code",
         "claude-web"
@@ -707,6 +893,7 @@
       "provider": "Anthropic",
       "tier": "claude.ai Max ($100)",
       "monthly_usd": 100.0,
+      "annual_usd": null,
       "surface_funded": [
         "claude-code",
         "claude-web"
@@ -717,6 +904,7 @@
       "provider": "Anthropic",
       "tier": "claude.ai Max ($200)",
       "monthly_usd": 200.0,
+      "annual_usd": null,
       "surface_funded": [
         "claude-code",
         "claude-web"
@@ -727,6 +915,7 @@
       "provider": "Cursor",
       "tier": "Cursor Pro",
       "monthly_usd": 20.0,
+      "annual_usd": 192.0,
       "surface_funded": [
         "cursor"
       ],
@@ -736,6 +925,7 @@
       "provider": "Cursor",
       "tier": "Cursor Pro+",
       "monthly_usd": 60.0,
+      "annual_usd": null,
       "surface_funded": [
         "cursor"
       ],
@@ -745,6 +935,7 @@
       "provider": "Cursor",
       "tier": "Cursor Ultra",
       "monthly_usd": 200.0,
+      "annual_usd": null,
       "surface_funded": [
         "cursor"
       ],
@@ -753,17 +944,19 @@
     {
       "provider": "Google",
       "tier": "Google AI Plus",
-      "monthly_usd": 7.99,
+      "monthly_usd": 4.99,
+      "annual_usd": null,
       "surface_funded": [
         "gemini-app",
         "gemini-cli"
       ],
-      "notes": "Entry-paid Google AI tier with 2x higher usage limits than Free in the Gemini app, access to video generation and Daily Brief, 200 Google Flow Credits, and 200 GB of cloud storage."
+      "notes": "Entry-paid Google AI tier with 2x higher usage limits than Free in the Gemini app, access to Gemini 3.1 Pro / Nano Banana Pro / Daily Brief / Gemini Omni video generation, 200 Google Flow Credits, and 400 GB of cloud storage (price cut from $7.99 to $4.99 on 2026-06-08; storage doubled from 200 GB to 400 GB)."
     },
     {
       "provider": "Google",
       "tier": "Google AI Pro",
       "monthly_usd": 20.0,
+      "annual_usd": 199.99,
       "surface_funded": [
         "gemini-app",
         "gemini-cli"
@@ -774,6 +967,7 @@
       "provider": "Google",
       "tier": "Google AI Ultra ($100)",
       "monthly_usd": 100.0,
+      "annual_usd": null,
       "surface_funded": [
         "gemini-app",
         "gemini-cli"
@@ -784,6 +978,7 @@
       "provider": "Google",
       "tier": "Google AI Ultra ($200)",
       "monthly_usd": 200.0,
+      "annual_usd": null,
       "surface_funded": [
         "gemini-app",
         "gemini-cli"
@@ -794,6 +989,7 @@
       "provider": "OpenAI",
       "tier": "ChatGPT Go",
       "monthly_usd": 8.0,
+      "annual_usd": null,
       "surface_funded": [
         "chatgpt-app",
         "codex-cli"
@@ -804,6 +1000,7 @@
       "provider": "OpenAI",
       "tier": "ChatGPT Plus",
       "monthly_usd": 20.0,
+      "annual_usd": null,
       "surface_funded": [
         "chatgpt-app",
         "codex-cli"
@@ -814,6 +1011,7 @@
       "provider": "OpenAI",
       "tier": "ChatGPT Pro ($100)",
       "monthly_usd": 100.0,
+      "annual_usd": null,
       "surface_funded": [
         "chatgpt-app",
         "codex-cli"
@@ -824,6 +1022,7 @@
       "provider": "OpenAI",
       "tier": "ChatGPT Pro ($200)",
       "monthly_usd": 200.0,
+      "annual_usd": null,
       "surface_funded": [
         "chatgpt-app",
         "codex-cli"