LLM hosting prices: what changed in May 2026
No price cuts and no rises.
0 price cuts · 0 price rises · 0 models newly sold per token · 0 withdrawn · 0 new GPU instances · 0 new regions
The clouds' own announcements
- Anthropic: We’ve deprecated fast mode for Claude Opus 4.6, with removal approximately 30 days after launch.
- Anthropic: We’ve launched mid-conversation system messages.
- Anthropic: We’ve launched Claude Opus 4.8 (claude-opus-4-8), our most capable model.
- Anthropic: We've deprecated fast mode for Claude Opus 4.6, with removal approximately 30 days after launch.
- Fireworks AI: Serverless migration guide
- Fireworks AI: Action required
- Anthropic: We’ve launched cache diagnostics in beta.
- Anthropic: We've launched cache diagnostics in beta.
- Azure: GPT Realtime Translate and GPT Realtime Whisper