2026/11/02
2026/10/19
2026/10/18
2026/10/13
- How much does speech-to-text cost per hour? Comparing the xAI and Mistral APIs
- Mistral Large 4 costs $1.36 in and $4.18 out, at half price for two weeks after launch
- Claude Managed Agents' web_fetch now fetches only URLs that already appear in the session
- xAI's speech-to-text API costs $0.10 an hour in batch, and moved to 2.0 on 2 October
- Claude Max and Team plans now include monthly API credits ($100 on Max 5x)
- GPT-6.1 Sol charges less than GPT-6 Sol for cached input
- Claude Sonnet 5.5 keeps Sonnet 5's price, and its cache reads were halved on 7 October
- Claude Haiku 5.5 is priced like GPT-6 Luna up to 100K tokens, and five times higher beyond
2026/09/29
2026/09/28
2026/09/15
- DeepSeek V4.1 Flash takes over from V4 Flash at a lower price
- Sakana AI's Fugu Ultra v2 and Fugu Max answer by routing work to other models
- Cognition's SWE-2 is a coding model built on Kimi K3
- Inception's Mercury 2.5, a diffusion model at $0.75 per million output tokens
- GPT-Live 1 bills by the minute, at $0.05, with the reasoning model on top
2026/09/08
- Meta sells the same model at two prices, and the cheap one is paid for with your prompts
- Grok 4.6 is the flagship, and it holds half the context of Grok 4.3
- Kimi K3 holds a million tokens at one flat rate
- The same Qwen model costs less from Beijing than from Singapore
- On Mistral, Large costs a third of Medium
- GPT-6 Astra and the three models that make up GPT-6
- GPT-6 Astra prices, and the 272K line that reprices the whole request
- Claude Fable 5.1 takes the top of the lineup, and Fable 5 moves to legacy
- Cache reads on Fable 5.1 are a quarter of what they cost on Fable 5
2026/09/02
- The current Claude lineup: 1M context, 128K output, retirement more than a year out
- SWE-bench Verified is at 97, and the top of the table is hard to read
- A benchmark number means nothing without the conditions it was taken under
- Claude API prices, and the Sonnet 5 increase that did not happen
- OpenAI API prices: the GPT-5.6 line has long-context rates too
- Gemini API prices, and the two dates that change them
- DeepSeek now charges double during peak hours
- Perplexity Sonar ends on 27 September 2026, replaced by the Agent API
- What prompt caching actually costs, and where the tenth does not hold
- Long context: who reprices the whole request, and who does not
- Adding an MCP server to Claude Code, and which scope to pick
