GPT-6.1 Sol vs Llama 4 Maverick: cost per task
Llama 4 Maverick is cheaper on 14 of 14 tasks and about 3.9x cheaper on average (geometric mean). GPT-6.1 Sol has the higher LMArena Text score (1484 vs 1327).
| Task | GPT-6.1 Sol | Llama 4 Maverick | Difference |
|---|---|---|---|
| Answer a chat question | 0.67¢ | 0.11¢ | 6.4x |
| Reply to a customer email | 0.18¢ | 0.05¢ | 3.5x |
| Write a 1,500-word blog post | 2.3¢ | 1.0¢ | 2.2x |
| Write 10 product descriptions | 1.2¢ | 1.1¢ | 1.2x |
| Translate 1,000 words | 1.6¢ | 0.11¢ | 14.5x |
| Proofread one page (500 words) | 0.81¢ | 0.13¢ | 6.0x |
| Summarize a 20-page PDF | 3.1¢ | 0.42¢ | 7.5x |
| Extract data from an invoice | 0.52¢ | 0.15¢ | 3.5x |
| Analyze a 200-row spreadsheet | 8.4¢ | 1.2¢ | 7.3x |
| Answer from 5 retrieved documents (RAG) | 1.2¢ | 0.26¢ | 4.6x |
| Classify 100 support tickets | 0.77¢ | 0.58¢ | 1.3x |
| Review 500 lines of code | 4.9¢ | 1.1¢ | 4.3x |
| Write a SQL query | 0.48¢ | 0.17¢ | 2.8x |
| Run a 50-step coding agent | 53¢ | 19¢ | 2.8x |
Specs and list prices
| GPT-6.1 Sol | Llama 4 Maverick | |
|---|---|---|
| Input per 1M | $2.00 | $0.19 |
| Output per 1M | $10.00 | $0.65 |
| Cached input per 1M | $0.10 | $0.05 |
| Context | 1.05M | 1M |
| Providers | 8 | 4 |
| Reasoning | always | none |
| LMArena Text | 1484 | 1327 |
| Released | Sep 29, 2026 | Apr 5, 2025 |
FAQ
Is GPT-6.1 Sol or Llama 4 Maverick cheaper?
Llama 4 Maverick is cheaper on 14 of 14 tasks and about 3.9x cheaper on average (geometric mean). GPT-6.1 Sol has the higher LMArena Text score (1484 vs 1327).
How much does GPT-6.1 Sol cost compared to Llama 4 Maverick per token?
GPT-6.1 Sol: $2.00 input and $10.00 output per 1M tokens. Llama 4 Maverick: $0.19 input and $0.65 output. Token prices alone mislead - models use different numbers of tokens for the same task, which is what the task table shows.