MetaLlama 3-70B
Llama 3-70B is a Meta open-weights model, released 18 Apr 2024. Its best configuration ranks #707 of 785 on the BenchLeader Index at 38.6 ±2.9, in the lower half. The ± is the point: 206 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in human preference (42) and lowest in long context (23). At $1.18 per million tokens blended it is pricier than most ranked models. Last measured 22 Oct 2026.
- Blended price
- $1.18/M
- $0.650 in · $2.75 out
- Output speed
- –
- First answer
- –
- Context
- 8k
- Full answer
- –
- Cost per run
- –
- Released
- 18 Apr 2024
- Overall index39
- Reasoning37
- Coding41
- Agents & tools35
- Maths36
- Knowledge37
- Instruction following40
- Human preference42
- Long context23
- Composite36
Versions
Meta has shipped 4 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Llama 3.3 70B | 6 Dec 2024 | 40.7 | #651 |
| Llama 3.1 70B | 23 Jul 2024 | 40.9 | #645 |
| Llama 3-70Bthis page | 18 Apr 2024 | 38.6 | #707 |
| Llama 2-70B | 18 Jul 2023 | 34.3 | #772 |
How its index has moved
8 Sept 2026 to 21 Oct 2026The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| GPQA Diamond | 40.6%#248 | Epoch AI Benchmarking Hub | |
| Arena Hard Prompts | 1280#275 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 37.9%#486 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 4.5%#497 | Artificial Analysis | |
| Kagi LLM Benchmark | 35.1%#117 | Kagi LLM Benchmark | |
| CritPt | 0.0%#315 | Artificial Analysis | |
| DTBenchnot in index | 54.2%#199 | Epoch AI Benchmarking Hub | |
| ForecastBenchnot in index | 60.2%#71 | ForecastBench | |
| WinoGrandenot in index | 83.5%#9 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Coding | 1306#274 | Arena (LMArena) |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Cybench | 5.0%#21 | Cybench | |
| Terminal-Bench Hard | 0.8%#349 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 0.0%#393 | Artificial Analysis |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| OTIS Mock AIME | 4.3%#249 | Epoch AI Benchmarking Hub | |
| MATH Level 5 | 22.6%#78 | Epoch AI Benchmarking Hub | |
| Arena Maths | 1258#264 | Arena (LMArena) |
Knowledge
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -54.3#427 | Artificial Analysis | |
| MMLUnot in index | 79.3%#23 | Epoch AI Benchmarking Hub | |
| AA-Omniscience: accuracynot in index | 17.7%#356 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 12.5%#373 | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| IFBench | 37.1%#293 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1259#274 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1277#273 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1254#262 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1273#265 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1254#288 | Arena (LMArena) |
Long context
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-LCR | 0.0%#493 | Artificial Analysis |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Epoch Capabilities Indexnot in index | 123.0#166 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index v4.3.2 | 5.5#580 | Artificial Analysis |
Price history
The price above, per 1M tokens, every day it changed.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0011 | – | – |
| Summarize a 30-page report | 12,000 / 600 | $0.0095 | – | – |
| Code edit | 6,000 / 1,500 | $0.0080 | – | – |
| Agentic coding session | 60,000 / 4,000 | $0.050 | – | – |
| Structured extraction | 2,000 / 200 | $0.0019 | – | – |
More from Meta
See also
Questions people ask
- What are Llama 3-70B's benchmark scores?
- As of 25 Oct 2026, Llama 3-70B scores 38.6 ±2.9 on the BenchLeader Index, strongest in human preference (42) and weakest in long context (23). That rests on 15 independent benchmark results, and every score on this page links to its publisher.
- How does Llama 3-70B rank?
- Llama 3-70B is #707 of 785 ranked configurations on 25 Oct 2026. 206 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Llama 3-70B cost per 1M tokens?
- As of 25 Oct 2026, Llama 3-70B costs $0.650 per 1M input tokens and $2.75 per 1M output tokens, $1.18 blended (three input tokens per output token). The list price last changed on 9 Sept 2026.
- When was Llama 3-70B released?
- Meta released Llama 3-70B on 18 Apr 2024.
Data as of 25 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Llama 3-70B: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/llama-3-70b.