MetaLlama 3.2 1B Instruct
Llama 3.2 1B Instruct is a Meta open-weights model, released 24 Sept 2024. Its best configuration ranks #779 of 785 on the BenchLeader Index at 32.4 ±3.3, in the lower half. The ± is the point: 52 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 knowledge (37) and lowest in human preference (22). At $0.071 per million tokens blended it is among the cheapest fifth of ranked models. Output speed of 127 tokens per second puts it faster than most, with a first token in 0.2 s. Last measured 22 Oct 2026.
- Blended price
- $0.071/M
- $0.027 in · $0.201 out
- Output speed
- 127 tok/s
- OpenRouter traffic, 7-day median; not yet measured by Artificial Analysis
- First answer
- 0.23 s
- Context
- 128k
- Full answer
- –
- Cost per run
- –
- Released
- 24 Sept 2024
- Overall index32
- Reasoning29
- Coding24
- Agents & tools28
- Maths26
- Knowledge37
- Instruction following28
- Human preference22
- Long context27
- Composite35
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 | 23.9%#286 | Epoch AI Benchmarking Hub | |
| Arena Hard Prompts | 1114#353 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 19.6%#556 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 5.5%#428 | Artificial Analysis | |
| CritPt | 0.0%#315 | Artificial Analysis |
Coding
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Coding | 1149#348 | Arena (LMArena) |
Agents & tools
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Terminal-Bench Hard | 0.0%#362 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 0.0%#393 | Artificial Analysis | |
| BFCL Overall | 10.8%#81 | Berkeley Function Calling Leaderboard | |
| BALROGnot in index | 6.6%#36 | BALROG |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| OTIS Mock AIME | 0.6%#271 | Epoch AI Benchmarking Hub | |
| Arena Maths | 1123#328 | Arena (LMArena) |
Knowledge
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -54.7#432 | Artificial Analysis | |
| AA-Omniscience: accuracynot in index | 7.0%#526 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 33.7%#198 | Artificial Analysis |
Instruction following
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| IFBench | 22.8%#403 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1086#354 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1111#356 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1082#353 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1076#347 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1102#345 | Arena (LMArena) |
Long context
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA-LCR | 6.7%#480 | Artificial Analysis |
Composite
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| AA Intelligence Index v4.3.2 | 4.8#618 | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Cloudflare | 143 tok/s | 0.22 s | $0.027 | $0.201 | 60k | – |
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 | $0.0001 | – | 2.6 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0004 | – | 4.9 s |
| Code edit | 6,000 / 1,500 | $0.0005 | – | 12.0 s |
| Agentic coding session | 60,000 / 4,000 | $0.0024 | – | 31.7 s |
| Structured extraction | 2,000 / 200 | $0.0001 | – | 1.8 s |
See also
Questions people ask
- What are Llama 3.2 1B Instruct's benchmark scores?
- As of 24 Oct 2026, Llama 3.2 1B Instruct scores 32.4 ±3.3 on the BenchLeader Index, strongest in knowledge (37) and weakest in human preference (22). That rests on 13 independent benchmark results, and every score on this page links to its publisher.
- How does Llama 3.2 1B Instruct rank?
- Llama 3.2 1B Instruct is #779 of 785 ranked configurations on 24 Oct 2026. 52 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 3.2 1B Instruct cost per 1M tokens?
- As of 24 Oct 2026, Llama 3.2 1B Instruct costs $0.027 per 1M input tokens and $0.201 per 1M output tokens, $0.071 blended (three input tokens per output token).
- When was Llama 3.2 1B Instruct released?
- Meta released Llama 3.2 1B Instruct on 24 Sept 2024.
Data as of 24 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Llama 3.2 1B Instruct: benchmarks, pricing, speed and rank”, data as of 24 Oct 2026, https://www.benchleader.com/models/llama-3-2-1b.