Which AI got dumber today?
Community ratings and official prices for 13 AI models, updated daily. Rate the ones you used today.
Dumb meter, last 24 hours
- Dumber
Lobotomized
Dumb
Meh
Solid
Big brain- Smarter
No votes yet today. Used one of these models? Tap a face, it takes one second.
- Claude Fable 5.1Anthropic–no votes yet
- Claude Opus 5.5Anthropic–no votes yet
- Claude Sonnet 5Anthropic–no votes yet
- Claude Haiku 4.5Anthropic–no votes yet
- GPT-6 AstraOpenAI–no votes yet
- GPT-6 SolOpenAI–no votes yet
- GPT-6 LunaOpenAI–no votes yet
- GPT-5.6 SolOpenAI–no votes yet
- GPT-5.6 LunaOpenAI–no votes yet
- Gemini 3.1 ProGoogle–no votes yet
- Gemini 3.8 FlashGoogle–no votes yet
- Grok 4.7xAI–no votes yet
- DeepSeek V4 ProDeepSeek–no votes yet
This week's ranking → Tap a face to rate a model you used today. Bar = share of Dumb + Lobotomized votes, face = average rating (from 5 votes).
Discussions
Start a discussion
What changed lately
- Claude Opus 5.5 releasedSep 22 · Release
- GPT-6 Sol and Luna releasedSep 22 · Release
- Grok 4.7 releasedSep 21 · Release
- DeepSeek V4.1 Flash releasedSep 10 · Release
- OpenAI status: elevated errors across ChatGPT and CodexSep 3 · Official incident
- GPT-6 Astra releasedSep 3 · Release
All news → · Prices re-checked September 28, 2026.
What they cost
List prices in USD per million tokens, cheapest first for a typical chat workload. Checked September 28, 2026.
| Model | Input /1M | Output /1M | 1,000 chats |
|---|---|---|---|
| GPT-6 Luna | $0.10 | $0.50 | $0.40 |
| GPT-5.6 Luna | $0.20 | $1.20 | $0.90 |
| Gemini 3.8 Flash | $0.75 | $3.75 | $3.00 |
| DeepSeek V4 Pro | $1.32 | $3.96 | $3.96 |
| Claude Haiku 4.5 | $1.00 | $5.00 | $4.00 |
| Grok 4.7 | $2.00 | $6.00 | $6.00 |
| Claude Sonnet 5 | $2.00 | $10.00 | $8.00 |
| GPT-6 Sol | $2.00 | $10.00 | $8.00 |
| Gemini 3.1 Propreview | $2.00 | $12.00 | $9.00 |
| Claude Opus 5.5 | $4.00 | $20.00 | $16.00 |
| GPT-5.6 Sol | $5.00 | $30.00 | $22.50 |
| Claude Fable 5.1 | $10.00 | $50.00 | $40.00 |
| GPT-6 Astra | $10.00 | $50.00 | $40.00 |
All text model prices →Image models →Video models →Cost calculator →Compare two models →
Why a model can feel dumber
Incidents
Providers post “degraded performance” notices and have confirmed quality bugs in public postmortems.
Changed defaults
Reasoning effort, routing and usage limits can change without a new model name.
Long chats
Research shows quality drops across long multi-turn conversations, whatever the model.