meta-llama
Meta: Llama 3.3 70B Instruct
meta-llama/llama-3.3-70b-instruct
Our take
AI-assisted analysis based on the sources below. Suitability is an editorial assessment; test it on your own workload.
Meta's Llama 3.3 70B Instruct is a 70B-parameter multilingual generative model that is pretrained and instruction-tuned, offered through OpenRouter's catalog as a text-in, text-out model. The listing reports a 131,072-token context window, a knowledge cutoff at the end of 2023, and support for tool calling, structured outputs, and common sampling controls. OpenRouter's catalog is the only evidence supplied; no independent model card was available in this source pack, and the pack contains only limited third-party coding metrics, so real-world performance should be validated on the target workload.[1]
Strengths
- According to the OpenRouter catalog, the model supports tool calling, tool choice, and structured outputs, alongside sampling controls such as temperature, top-p, min-p, and logit bias.[1]
- The listing reports a 131,072-token context length, which may suit long-document or multi-turn applications.[1]
- The catalog describes it as a multilingual instruction-tuned generative model, suggesting suitability for non-English workloads.[1]
Limitations
- All evidence comes from a single OpenRouter catalog entry; no independent model card or official documentation was available in this source pack, so capability and quality claims remain unverified.[1]
- The listing reports text-only input and output; no image, audio, or other modalities are documented.[1]
- The catalog does not report a dedicated reasoning mode, and the pack contains only one third-party coding metric, so measured performance is largely unknown and needs workload validation.[1]
Where to use it
- Agentic or tool-using assistants could be a reasonable fit given the catalog-reported function calling and structured output support; test against your own tools before adopting.[1]
- Multilingual chat or drafting workloads may benefit from the reported multilingual instruction tuning, though per-language quality is unmeasured in this pack.[1]
- Long-context tasks such as document Q&A could leverage the listed 131k-token context window, pending validation of recall and latency.[1]
Sources
- [1] OpenRouter catalog · Meta: Llama 3.3 70B InstructSources checked: Nov 13, 2026
About this model
Description below is the original text supplied by OpenRouter, shown in the source language.
The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model...
Specifications
- Model ID
- meta-llama/llama-3.3-70b-instruct
- Creator
- meta-llama
- Added to OpenRouter
- 06.12.2024
- Context length
- 131,072
- Modality
- text->text
- Input modalities
- text
- Output modalities
- text
- Token pricing
- USD per 1M tokens: in $0.710 / out $0.710
- Knowledge cutoff
- 31.12.2023
- Expiration date
- —
Tasks and capabilities
A grade appears only where the source supplies a matching measurement for that task, the model is active and the cohort holds at least 10 comparable models. The capability column lists facts the source declares — supported formats, reasoning and tool parameters — which are not measured quality.
| Task | Grade | Source-declared capability |
|---|---|---|
| General quality | No measurement | — |
| Coding | CC — below the B band | — |
| Agents | No measurement | parameters: structured_outputs, tool_choice, tools |
| Writing | No measurement | output: text |
| Reasoning | No measurement | — |
| Document analysis | No measurement | — |
| Vision | No measurement | — |
| Image generation | No measurement | — |
| Audio | No measurement | — |
| Video | No measurement | — |
| Embeddings | No measurement | — |
Evaluations
| Source | Benchmark | Value | Grade | Data retrieved |
|---|---|---|---|---|
| Artificial Analysis | Coding index | 11.9 | CC — below the B band | 13.11.2026 |
How grades work
Each letter grade compares a model only against other models scored by the same source and metric in the current snapshot. S means the top 10%, A the next 20%, B the next 30%, C the rest. Models without a published measurement are shown as not rated — a supported format alone never implies measured quality.
- S — S — top 10% of the cohort
- A — A — next 20% of the cohort
- B — B — next 30% of the cohort
- C — C — below the B band
Providers
Optional live provider statistics from OpenRouter. Missing values are unknown, not zero.
| Provider | Context length | Price in /1M | Price out /1M | Latency p50 (30m, s) | Throughput p50 (30m, tok/s) | Uptime (30m) |
|---|---|---|---|---|---|---|
| DeepInfrafp8 | 131,072 | $0.100 | $0.320 | — | — | 100% |
| AkashMLfp8 | 131,072 | $0.200 | $0.520 | — | — | 100% |
| Parasailfp8 | 131,072 | $0.220 | $0.500 | — | — | 100% |
| Cloudflarefp8 | 24,000 | $0.293 | $2.25 | — | — | 95.65% |
| Groq | 131,072 | $0.590 | $0.790 | — | — | 100% |
| SambaNova | 131,072 | $0.600 | $1.2 | — | — | 100% |
| CoreWeavefp16 | 128,000 | $0.710 | $0.710 | — | — | 100% |
| 128,000 | $0.720 | $0.720 | — | — | — | |
| 128,000 | $0.720 | $0.720 | — | — | — |
Provider data retrieved: 13.11.2026
Related articles
No published articles identify this model yet.