Compare / Llama 3.3 70B Instruct vs Llama 4 Maverick
Llama 3.3 70B Instruct vs Llama 4 Maverick
Which is cheaper, which is more capable, and which does better at each kind of work, from 20 results on 3 sources that measured both models.
Llama 3.3 70B Instruct or Llama 4 Maverick? The short answer
- Llama 4 Maverick is cheaper: $0.42 against $0.64 per million tokens, 35% less.
- Llama 4 Maverick scores higher on the Artificial Analysis Intelligence Index: 10.0 against 7.7.
- Llama 3.3 70B Instruct writes faster: 80 tokens a second against 42.
- On our business benchmarks, Llama 3.3 70B Instruct ranks higher for agents and tool use and writing.
- On our business benchmarks, Llama 4 Maverick ranks higher for reasoning and knowledge.
At a glance
| Llama 3.3 70B Instruct | Llama 4 Maverick | |
|---|---|---|
| Price per million tokens | $0.59 in · $0.79 out | $0.27 in · $0.85 out |
| Blended (3 in : 1 out) | $0.64 | $0.42 |
| Cheapest host, blended | $0.16 (DeepInfra, FP8) | $0.30 (DigitalOcean) |
| Intelligence Index | 7.7 | 10.0 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | 80 | 42 |
| Context window | 131,072 tokens | 1,048,576 tokens |
| Weights | Open | 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 | Llama 3.3 70B Instruct | Llama 4 Maverick |
|---|---|---|
| Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls | 127 of 161 | 141 of 161 |
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 174 of 195 | 146 of 195 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | 165 of 187 | 166 of 187 |
Every shared result
| Benchmark | Metric | Llama 3.3 70B Instruct | Llama 4 Maverick |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,304 | 1,317 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,285 | 1,307 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,298 | 1,326 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,293 | 1,314 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,318 | 1,327 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,292 | 1,312 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 11.9 | 16.3 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 7.70 | 10.0 |
| Artificial Analysisreported | GPQA Diamond% of questions | 49.8% | 67.1% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 3.6% | 4.9% |
| Artificial Analysisreported | IFBench% of instructions | 47.1% | 43.0% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 15.7% | 50.0% |
| Artificial Analysisreported | Output speedtokens per second | 80.5 | 42.4 |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 4.9% | 7.9% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 3.0% | 6.8% |
| Artificial Analysisreported | Time to first answer tokenseconds | 0.64 | 0.64 |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 26.6% | 17.8% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 10.0% | 13.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.37 | 0.37 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 26.2 | 24.0 |
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 Llama 3.3 70B Instruct better than Llama 4 Maverick?
It depends on the task. Among models available today, on our benchmarks and licensed sources, Llama 3.3 70B Instruct ranks higher for agents and tool use and writing; Llama 4 Maverick ranks higher for reasoning and knowledge.
Which is cheaper, Llama 3.3 70B Instruct or Llama 4 Maverick?
Llama 4 Maverick costs $0.42 per million tokens (three input to one output), against $0.64 for Llama 3.3 70B Instruct.
Which is faster, Llama 3.3 70B Instruct or Llama 4 Maverick?
Llama 3.3 70B Instruct writes about 80 tokens a second on its usual API, against 42 for Llama 4 Maverick (Artificial Analysis).
Run Llama 3.3 70B Instruct and Llama 4 Maverick on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.