Compare / Claude Opus 4.6 vs Llama 3.1 70B Instruct
Claude Opus 4.6 vs Llama 3.1 70B Instruct
Which is cheaper, which is more capable, and which does better at each kind of work, from 15 results on 3 sources that measured both models.
Claude Opus 4.6 or Llama 3.1 70B Instruct? The short answer
- Llama 3.1 70B Instruct is cheaper: $0.40 against $10.00 per million tokens, 25 times less.
- Claude Opus 4.6 scores higher on the Artificial Analysis Intelligence Index: 31.9 against 6.6.
- On our business benchmarks, Claude Opus 4.6 ranks higher for reasoning and knowledge and writing.
At a glance
| Claude Opus 4.6 | Llama 3.1 70B Instruct | |
|---|---|---|
| Price per million tokens | $5.00 in · $25.00 out | $0.40 in · $0.40 out |
| Blended (3 in : 1 out) | $10.00 | $0.40 |
| Cheapest host, blended | $10.00 (Claude Platform on AWS) | $0.40 (DeepInfra, FP8) |
| Intelligence Index | 31.9 | 6.6 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | – | – |
| Context window | 1,000,000 tokens | 131,072 tokens |
| Weights | Closed | Open |
| Developer based in | the US | the US |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.
| Use case | Claude Opus 4.6 | Llama 3.1 70B Instruct |
|---|---|---|
| Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls | 17 of 161 | not measured |
| Professional workReal tasks from banking, consulting and law, business documents and freelance projects | 27 of 51 | not measured |
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 45 of 195 | 179 of 195 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | 3 of 187 | 153 of 187 |
| Sticking to the factsSummarising without inventing things, and factual answers | 57 of 153 | not measured |
Every shared result
| Benchmark | Metric | Claude Opus 4.6 | Llama 3.1 70B Instruct |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,501 | 1,287 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,501 | 1,257 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,547 | 1,275 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,513 | 1,273 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,505 | 1,293 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,500 | 1,264 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 31.9 | 6.60 |
| Artificial Analysisreported | GPQA Diamond% of questions | 89.6% | 40.9% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 39.9% | 4.5% |
| Artificial Analysisreported | IFBench% of instructions | 53.1% | 34.4% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 48.5% | 3.0% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 92.1% | 15.2% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 3.0% | 15.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.41 | 0.38 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 70.9 | 32.8 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.
Quick answers
Is Claude Opus 4.6 better than Llama 3.1 70B Instruct?
It depends on the task. Among models available today, on our benchmarks and licensed sources, Claude Opus 4.6 ranks higher for reasoning and knowledge and writing.
Which is cheaper, Claude Opus 4.6 or Llama 3.1 70B Instruct?
Llama 3.1 70B Instruct costs $0.40 per million tokens (three input to one output), against $10.00 for Claude Opus 4.6.
Run Claude Opus 4.6 and Llama 3.1 70B Instruct on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.