AI MODELS
133 items
132 posts, 1 guide
Qwen4-Max is live on QwenCloud at $1.50 input and $4.50 output per million tokens, down from $2 and $6 for Qwen3.8-Max. What is vendor-reported, what is missing, and a test plan.
Amazon engineers replaced Sonnet 4.6 on a 20-label ticket classifier with Strands Decider, a small decision model, at about 44x lower list cost. Here is the narrow job it fits and the rollout pattern worth copying.
Kimi K3.1 Pro costs $2/$10 per million tokens, promises weights by Nov 4, and scores 65.7% on Terminal-Bench 4.0 (vendor-reported). The cost-per-task catch.
Mistral Large 4 weights are on Hugging Face under a modified MIT license. What the license blocks, the vLLM command, and when self-hosting beats the API.
Anthropic filed its S-1 for a Nasdaq listing under ANT. The revenue and loss figures, the risk factors that touch API users, and what to change in your stack on Monday.
Recursive's Flint is a 40B open-weight coding model with 2.5B active parameters. How to run it on Ollama and point Claude Code at it, plus the benchmark caveats.
Every scheduled AI model API shutdown through mid 2027 by month: GPT-4o on Jan 14, whisper-1 on Feb 26, Claude Sonnet 4.5 on Nov 30, plus how to migrate.
97 of 167 models we track enforce a JSON schema on output. Request shapes, schema limits and refusal handling for OpenAI, Claude, Gemini and Mistral.
DeepSeek V4.1-Flash-1017 jumps 11 points on Terminal-Bench 4.0 at the same deepseek-flash price. Benchmarks vs Opus 5, K3, GLM-5.3 and when to switch.
Codex says gpt-5.5 is not supported with a ChatGPT account because OpenAI retired it from ChatGPT sign-in on October 14. The fix and what still works.
LFM2.5-24B-A2B vs Qwen3.6-35B-A3B for local agents: Liquid wins tool-use benchmarks, Qwen wins SWE-Bench Verified. Which one to run for which job.
Ollama 0.40.4 runs decision models locally: ollama run nimble --ask answers yes/no, label and score questions with probabilities. Commands and limits.
DeepSeek 4.1 Flash as a daily coding model: what a session costs, how peak hours double the bill, why caching matters, and where the subsidy argument holds.
Which cheap model belongs under your orchestrator? Haiku 5.5 vs GPT-6 Luna as a subagent or worker: routing by prompt length, compaction, tool-call gotchas and cost per agent task.
OpenAI Decisions API costs $0.10 per million input tokens, 2.4x Jev. Pick Jev for cheap text decisions, Clef-flash for speed, Decisions API for images.
Step 5 Preview is StepFun's 600B MoE with 27B active and a 1M window. What it scores, what a session costs, and what to check before the October 15 weights.
Sonnet 5.5 is the default, Opus 5.5 is for open-ended long-horizon work, and Haiku 5.5 is for bounded steps and subagents. Prices, benchmarks, cost math.
Claude Haiku 5.5 pricing, the cliff above 100K-token prompts, the tokenizer catch, the migration steps from Haiku 4.5, and when to use it instead of Sonnet.

Get Smarter About AI Dev
New tutorials, open-source projects, and deep dives on coding agents - delivered weekly.