Releases·Tool Release
cline/cline CLI v3.0.66: Bun 1.4.2 builds, Windows proxy fix, and refreshed default model catalog
Cline shipped CLI v3.0.66 on GitHub, rebuilding binaries with Bun 1.4.2 so macOS 27 no longer kills them at launch and x64 builds run without AVX2. The release fixes Windows system-proxy hub startup, Esc-to-stop handling, Bedrock routing for GPT-6/GPT-5.6 and Nova models, and refreshes default models across many providers to GPT-6.1 Sol, Claude Sonnet 5.5, and others. Practitioners using unpinned models should expect changed defaults.
0 engagement1 sourceWed, Sep 30, 2026, 02:41 AM
AI Overview
- CLI binaries now build with Bun 1.4.2, fixing macOS 27 launch kills and letting x64 builds run on CPUs without AVX2
- Hub now starts on Windows behind system proxies like Clash and v2ray; loopback discovery requests no longer route through the proxy
- Esc right after sending a prompt now stops the turn instead of being ignored while setup was still in progress
- Content-filter blocks now explain the block and suggest rephrasing instead of the misleading "Model returned empty response"
- Anthropic server-side refusal fallback lets Claude models on OpenRouter and Cline fail over to another upstream provider
- Reasoning tokens no longer count twice in output totals; cost is unchanged since reasoning bills at the output rate
- Bedrock fixes: GPT-6 and GPT-5.6 route through inference profiles, India regions resolve the in. profile, and Nova 2 Lite requests stop being rejected
- Model catalog refresh adds Bee and Pareto providers, recommends Claude Sonnet 5.5 and Opus 5.5, and makes GPT-6.1 Sol the default across OpenAI, OpenRouter, Copilot, and others
Related
Releases·Tool Release
Cline Desktop v0.0.38 adds Linux ARM support and defaults OpenAI providers to GPT-6.1 Sol
Cline shipped Desktop v0.0.38 on GitHub, adding Linux ARM64 packages, cloud session persistence, and a range of reliability fixes. The release also refreshes the default model catalog across providers, making GPT-6.1 Sol the default for OpenAI and most gateways.
Cline shipped Desktop v0.0.38 on GitHub, adding Linux ARM64 packages, cloud session persistence, and a range of reliability fixes. The release also refreshes the default model catalog across providers, making GPT-6.1 Sol the default for OpenAI and most gateways.
0 engagement1 sourcegithubWed, Sep 30, 2026, 03:16 AM
Releases·Tool Release
cline/cline v4.1.22
Cline shipped v4.1.22 on GitHub, adding Bee (by HEOSSI) and Pareto Inference as providers and enabling Anthropic server-side refusal fallback so Claude models can fail over to another upstream provider on OpenRouter and Cline. It makes GPT-6.1 Sol the default model across ten OpenAI-compatible providers, refreshes the recommended list to Claude Sonnet 5.5 and Claude Opus 5.5, and fixes cancel, checkpoint, and token-counting bugs. Practitioners using non-pinned defaults should expect different models after update.
Cline shipped v4.1.22 on GitHub, adding Bee (by HEOSSI) and Pareto Inference as providers and enabling Anthropic server-side refusal fallback so Claude models can fail over to another upstream provider on OpenRouter and Cline. It makes GPT-6.1 Sol the default model across ten OpenAI-compatible providers, refreshes the recommended list to Claude Sonnet 5.5 and Claude Opus 5.5, and fixes cancel, checkpoint, and token-counting bugs. Practitioners using non-pinned defaults should expect different models after update.
0 engagement1 sourcegithubWed, Sep 30, 2026, 04:10 AM
Social·Community
Opus 5.5 vs Sonnet 5.5: 3D steampunk whale modeling
Users benchmarked Claude Sonnet 5.5 on 3D generation tasks against Opus 5.5 and OpenAI's GPT 6.1 Sol using identical prompt workflows. Sonnet 5.5 costs less and impressed on a Blender/three.js steampunk whale scene but trails Opus 5.5; on a three.js keyboard animation, GPT 6.1 Sol finished ~11 minutes for $1.78 API-equivalent versus Sonnet's 78 minutes at $57.34 — roughly 30x the cost. Both tests ran in atomic.chat.
Users benchmarked Claude Sonnet 5.5 on 3D generation tasks against Opus 5.5 and OpenAI's GPT 6.1 Sol using identical prompt workflows. Sonnet 5.5 costs less and impressed on a Blender/three.js steampunk whale scene but trails Opus 5.5; on a three.js keyboard animation, GPT 6.1 Sol finished ~11 minutes for $1.78 API-equivalent versus Sonnet's 78 minutes at $57.34 — roughly 30x the cost. Both tests ran in atomic.chat.
Updated Wed, Sep 30, 2026, 03:24 PM — A second user benchmark found GPT 6.1 Sol matches Sonnet 5.5 on a three.js task at ~1/30th the API-equivalent cost.
1.6k engagement2 sourcesredditTue, Sep 29, 2026, 12:54 PM
Releases·Model Releasefeatured
Anthropic releases Sonnet 5.5, a cheaper, 30% faster mid-tier model with Opus-level cyber capabilities
Anthropic released Claude Sonnet 5.5, the second model in the Claude 5.5 family, positioning it as a faster, lower-cost complement to Opus 5.5 for well-scoped coding and knowledge work. The company says it runs over 30% faster than Sonnet 5 and effectively costs up to 30% less per task despite identical per-token pricing, and it is the first Sonnet subject to the cybersecurity safeguards applied to Anthropic's most capable models. It is available everywhere, including Amazon Bedrock, and Haiku 5.5 is expected in the coming weeks.
Anthropic released Claude Sonnet 5.5, the second model in the Claude 5.5 family, positioning it as a faster, lower-cost complement to Opus 5.5 for well-scoped coding and knowledge work. The company says it runs over 30% faster than Sonnet 5 and effectively costs up to 30% less per task despite identical per-token pricing, and it is the first Sonnet subject to the cybersecurity safeguards applied to Anthropic's most capable models. It is available everywhere, including Amazon Bedrock, and Haiku 5.5 is expected in the coming weeks.
23022.0k engagement5 sourcesreddit, rss, xMon, Sep 28, 2026, 06:00 PM
Releases·Tool Release
cline CLI v3.0.57 adds hub drain and upgrade commands
cline/cline CLI v3.0.57 introduces `cline hub drain` and `cline hub upgrade` for safer hub maintenance, plus session persistence across restarts. The update also fixes tool calling being silently disabled for custom OpenAI-compatible endpoints.
cline/cline CLI v3.0.57 introduces `cline hub drain` and `cline hub upgrade` for safer hub maintenance, plus session persistence across restarts. The update also fixes tool calling being silently disabled for custom OpenAI-compatible endpoints.
0 engagement1 sourcegithubSun, Aug 23, 2026, 12:04 AM