pi-free 2.1.0 → 2.2.0

package/CHANGELOG.md

@@ -7,6 +7,51 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [2.2.0] - 2026-06-19
+
+### Added
+
+- **b.ai** provider (`api.b.ai/v1`) — OpenAI-compatible gateway with 29 models
+  (Claude Opus 4.x / Sonnet 4.x, GPT-5.x family, Gemini 3.x family, DeepSeek
+  V4 / V3.2, GLM-5.x, Kimi K2.5, Qwen 3.6-27B, MiniMax M3 / M2.7). Currently
+  advertises `MiniMax-M3` as a limited-time free promotional model; all
+  other models default to paid (visible by default). Static API key auth via
+  `BAI_API_KEY` env var or `bai_api_key` in `~/.pi/free.json`. New commands:
+  `/toggle-bai`.
+
+### Security
+
+- Bumped `@earendil-works/pi-ai`, `@earendil-works/pi-coding-agent`, and
+  `@earendil-works/pi-tui` to `^0.79.8` — resolves Dependabot advisories
+  #19 and #21 (high-severity protobufjs, ws, undici vulnerabilities in
+  `@earendil-works/pi-coding-agent`).
+- Bumped `typescript` devDep to `^6.0.3`.
+- Removed now-redundant `overrides` for `protobufjs` and `ws` (fixed
+  upstream).
+
+## [2.1.1] - 2026-06-15
+
+### Fixed
+
+- **Cline XML bridge**:
+  - Preserve JSON file content as string in `write_to_file` XML — prevents file bodies from being parsed as JSON objects and corrupted ([#244](https://github.com/apmantza/pi-free/pull/244)).
+  - Recover heredoc file writes (Model `cat << 'EOF'` pattern in `execute_command`) as `write`/`write_to_file` tool calls ([#246](https://github.com/apmantza/pi-free/pull/246)).
+  - Recover XML tool calls from the reasoning stream when MiMo nests tools inside thinking blocks ([#249](https://github.com/apmantza/pi-free/pull/249)).
+  - Surface reasoning-only responses: when MiMo puts the entire answer in reasoning with no visible text, surface it as best-effort visible output instead of a blank stop ([#251](https://github.com/apmantza/pi-free/pull/251)).
+  - Strip Unicode math-italic XML tag decorations (`<𝑎𝑛𝑡𝑚𝑙:thinking>`, `<𝑎𝑛𝑡𝑚𝑙:read_file>`) that MiMo emits instead of standard Cline XML tags ([#252](https://github.com/apmantza/pi-free/pull/252)).
+  - Hide internal planning phrases and restrict hidden-tool recovery to the reasoning channel only — never leak raw LLM planning as user-visible text ([#252](https://github.com/apmantza/pi-free/pull/252)).
+  - Retry MiMo stream errors with reasoning disabled on the second attempt ([#252](https://github.com/apmantza/pi-free/pull/252)).
+  - Parse MiMo Pi SDK `<function=name>` tool-call syntax directly — no double conversion through Cline XML ([#255](https://github.com/apmantza/pi-free/pull/255)).
+  - Auto-retry reasoning-only MiMo responses with a "continue" nudge instead of showing a dead-end error to the user ([#256](https://github.com/apmantza/pi-free/pull/256)).
+
+- **TokenRouter**:
+  - Patch nested MiniMax `<think>` blocks that appear inside `reasoning_content` deltas ([#247](https://github.com/apmantza/pi-free/pull/247)).
+  - Scope MiniMax thinking patches to active MiniMax models only, avoiding interference with other model families ([#248](https://github.com/apmantza/pi-free/pull/248)).
+  - Patch MiniMax payloads in the stream wrapper to prevent malformed SSE from breaking the parser ([#250](https://github.com/apmantza/pi-free/pull/250)).
+  - Retry high-load 2064 errors from TokenRouter with automatic backoff ([#254](https://github.com/apmantza/pi-free/pull/254)).
+
+- **UI**: Remove provider-count footer status text unconditionally — reduces status bar clutter ([#245](https://github.com/apmantza/pi-free/pull/245)).
+
 ## [2.1.0] - 2026-06-15
 
 ### Added
@@ -349,9 +394,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   - If ANY model has cost > 0, assumes pricing exposed → uses Route A
   - All providers (Cline, Kilo, NVIDIA, Ollama, dynamic built-in) now use this consistent helper
 
-- **CrofAI provider (PAID)** — Added new **paid** provider for CrofAI (https://crof.ai), an OpenAI-compatible LLM inference API. **Note: CrofAI is a paid provider** — users must have a CrofAI API key with credits. The provider uses Route B detection (name-only) since CrofAI's API doesn't expose per-model pricing. Only models with `"free"` in their names are marked as free (none currently).
+- **CrofAI provider (PAID)** — Added new **paid** provider for CrofAI (<https://crof.ai>), an OpenAI-compatible LLM inference API. **Note: CrofAI is a paid provider** — users must have a CrofAI API key with credits. The provider uses Route B detection (name-only) since CrofAI's API doesn't expose per-model pricing. Only models with `"free"` in their names are marked as free (none currently).
 
-- **ZenMux provider (PAID)** — Added new **paid** provider for ZenMux AI gateway (https://zenmux.ai), a unified API for 200+ models from OpenAI, Anthropic, Google, etc. **Note: ZenMux is a paid provider** — users must have a ZenMux API key with credits. The provider uses Route A detection (OR logic) since ZenMux exposes pricing. Models marked as free only if `cost === 0` OR `"free"` in name (2 free models identified: GLM 4.7 Flash Free, GLM 4.6v Flash Free).
+- **ZenMux provider (PAID)** — Added new **paid** provider for ZenMux AI gateway (<https://zenmux.ai>), a unified API for 200+ models from OpenAI, Anthropic, Google, etc. **Note: ZenMux is a paid provider** — users must have a ZenMux API key with credits. The provider uses Route A detection (OR logic) since ZenMux exposes pricing. Models marked as free only if `cost === 0` OR `"free"` in name (2 free models identified: GLM 4.7 Flash Free, GLM 4.6v Flash Free).
 
 - **Comprehensive `isFreeModel` test suite** — Added 30+ unit tests covering Route A, Route B, freemium behavior, and edge cases. Tests verify correct classification on actual OpenRouter API data (371 models, 30 free).
 
@@ -553,7 +598,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   - **$0.011 per 1,000 Neurons** beyond free allocation
   - Only requires `CLOUDFLARE_API_TOKEN` — account ID auto-derived from token
   - Toggle with `/cloudflare-toggle`
-  - Create token at https://dash.cloudflare.com/profile/api-tokens
+  - Create token at <https://dash.cloudflare.com/profile/api-tokens>
 
 - **Unified dynamic built-in providers module** — New `providers/dynamic-built-in/` module that dynamically fetches models from Pi's built-in providers when users have API keys:
   - **Mistral** (`MISTRAL_API_KEY`) — Fetches from `api.mistral.ai/v1/models`

package/README.md

@@ -14,7 +14,7 @@ Free and paid AI model providers for [Pi](https://pi.dev). Access **free and pai
 
 When you install pi-free, it:
 
-1. **Registers free-tier providers** with Pi's model picker — Kilo (free), Cline (free), LLM7 (free), TokenRouter (1 free model), ZenMux (paid), CrofAI (paid), Ollama Cloud (freemium), SambaNova (freemium), Codestral (freemium), DeepInfra (trial credit), Together AI (trial credit), Novita (paid), Routeway (paid), and more
+1. **Registers free-tier providers** with Pi's model picker — Kilo (free), Cline (free), LLM7 (free), TokenRouter (1 free model), ZenMux (paid), CrofAI (paid), b.ai (paid, 1 free promo model), Ollama Cloud (freemium), SambaNova (freemium), Codestral (freemium), DeepInfra (trial credit), Together AI (trial credit), Novita (paid), Routeway (paid), and more
 
 2. **Captures Pi's built-in OpenCode and OpenRouter providers** with a free/paid toggle — OpenCode and OpenRouter are now built into Pi; pi-free adds `/toggle-opencode` and `/toggle-openrouter` so you can switch between free-only and all models without restart
 
@@ -341,7 +341,7 @@ Codestral is Mistral's code-focused model via `codestral.mistral.ai`:
 
 - Free tier (Experiment plan): 2 req/min, 500K tokens/min, 1B tokens/month
 - No credit card — phone verification only
-- Sign up at https://console.mistral.ai/codestral
+- Sign up at <https://console.mistral.ai/codestral>
 
 ```bash
 export CODESTRAL_API_KEY="..."
@@ -363,7 +363,7 @@ LLM7 routes across multiple providers through a single OpenAI-compatible endpoin
 
 - Free tier: default/fast selectors, 100 req/hr, 20 req/min
 - No credit card required
-- Get free token at https://token.llm7.io/
+- Get free token at <https://token.llm7.io/>
 
 ```bash
 export LLM7_API_KEY="..."
@@ -404,6 +404,34 @@ Fast inference on custom RDU hardware:
 export SAMBANOVA_API_KEY="..."
 ```
 
+### b.ai (paid gateway with 1 free promo model)
+
+OpenAI-compatible gateway at `api.b.ai/v1` providing access to 29 frontier models:
+
+- **Anthropic**: Claude Opus 4.5–4.8, Claude Sonnet 4.5–4.6, Claude Haiku 4.5
+- **OpenAI**: GPT-5.5, GPT-5.5 Instant, GPT-5.4, GPT-5.4 Pro/Mini/Nano, GPT-5.2, GPT-5 Mini/Nano
+- **Google**: Gemini 3.1 Pro, Gemini 3 Flash, Gemini 3.5 Flash
+- **DeepSeek**: V4 Pro, V4 Flash, V3.2
+- **GLM**: GLM-5.x family
+- **Other**: Kimi K2.5, Qwen 3.6-27B, MiniMax M3 / M2.7
+
+Currently advertises `MiniMax-M3` as a limited-time free promotional model. All other models are paid — they're visible by default (toggle with `/toggle-bai` to show free-only).
+
+```bash
+export BAI_API_KEY="sk-..."
+```
+
+Or add to `~/.pi/free.json`:
+
+```json
+{
+  "bai_api_key": "sk-...",
+  "bai_show_paid": true
+}
+```
+
+Get a key at <https://b.ai/>.
+
 ---
 
 ## Slash Commands
@@ -426,6 +454,7 @@ Each provider has toggle commands to switch between free and all models:
 | `/toggle-deepinfra`     | Toggle DeepInfra (💳 trial credit)                       |
 | `/toggle-together`      | Toggle Together AI (💳 trial credit)                     |
 | `/toggle-sambanova`     | Toggle SambaNova (🔄 freemium)                           |
+| `/toggle-bai`           | Toggle b.ai (💳 paid — 1 free promo model)                |
 | `/toggle-llm7`          | Toggle LLM7 (✅ free gateway)                            |
 | `/toggle-zenmux`        | Toggle ZenMux (💳 paid)                                  |
 | `/toggle-crofai`        | Toggle CrofAI (💳 paid)                                  |
@@ -443,7 +472,6 @@ Each provider has toggle commands to switch between free and all models:
 - **For 🔧 dynamic API providers**: Filters the model list when you have an API key configured
 - **Persists your preference** to `~/.pi/free.json` for next startup
 
-
 ### Probe Commands (Health Check)
 
 Test models for errors and auto-hide broken ones:

package/banner.svg

@@ -96,7 +96,7 @@
     <text x="20" y="70" font-family="system-ui, sans-serif" font-size="11" fill="#8b9aaa">Ollama Cloud · SambaNova</text>
     <text x="20" y="90" font-family="system-ui, sans-serif" font-size="11" fill="#8b9aaa">Codestral · LLM7 · Novita</text>
     <text x="20" y="110" font-family="system-ui, sans-serif" font-size="11" fill="#8b9aaa">ZenMux · CrofAI · DeepInfra</text>
-    <text x="20" y="130" font-family="system-ui, sans-serif" font-size="11" fill="#8b9aaa">Together · TokenRouter · FastRouter</text>
+    <text x="20" y="130" font-family="system-ui, sans-serif" font-size="11" fill="#8b9aaa">Together · TokenRouter · b.ai</text>
   </g>
 
   <g transform="translate(970, 55)">