Model updates
18 articles
Launches and evaluations from this edition, newest articles first.
No model updates match these filters.
Evaluation
GPT-6.1 Sol vs Astra: move the testable work, keep the hard calls
On pirouter
GPT-6.1 Sol on pirouter: $2/M input and a listed long-context tier
On pirouter
Claude Sonnet 5.5 on pirouter: exact model ID and $2/M standard input
Evaluation
Claude Sonnet 5.5 task cost: why 30% less and 50% more describe different tests
Evaluation
What breaks when you switch to Claude Opus 5.5: four 400s, three silent changes, and what your fallback loses
Evaluation
GPT-6 Sol vs Claude Opus 5.5 cost per task: Sol costs 28% to 79% as much for the same target score
Evaluation
What a 250K prompt costs on Grok 4.7, GPT-6 Sol and Opus 5.5: the 200K and 272K price cliffs
Vendor release
DeepSeek V4.1 Flash price cut: 9% to 57% depending on your token mix, and the deepseek-v4-pro switch DeepSeek later reversed
Evaluation
Is GPT-6 Astra worth $10/$50? Two evals disagree, and both are right
Evaluation
GPT-6 Astra vs Claude Fable 5.1: same $10/$50, and a 39% cheaper agent bill
Evaluation
Is Claude Fable 5.1 cheaper than Opus 5? Only when cache reads are a third of your bill
Vendor release
Astra's "Critical" cyber threshold, Mythos 5.1 and Gemini 3.8 Flash Cyber: one model, two safety tiers, and API access is now a spec-sheet line
Evaluation
Why your IQ2 GGUF is the same size as IQ4: the 256-block fallback in llama-quantize
Vendor release
GLM-5.3 open weights, day one: 16 providers, five prices, four quantizations — and BF16 is the cheapest
Vendor release
Kimi K3 license, commercial use: the $20M MaaS clause, the 100M-MAU badge, and the 30% Moonshot wants from Azure, AWS and Google
Evaluation
FP8 flips 20% of top-1 tokens, and tool calls break first
Evaluation
Qwen3.8-27B: seven endpoints, one name, two prices
Vendor release
ox-alpha vs deepseek-v4-flash: what a stealth model hides