Llama 3.3 70B
Llama 3.3 Instruct 70B
Intelligence
7.66
#410 of 520 models
Output speed
86.1 tok/s
Measured generation rate
Cost / index task
—
Measured evaluation workload
End-to-end latency
7ms
First answer 2ms
01 / Position
A model in the landscape
Compare this model with peers across intelligence, speed, cost, and individual evaluations.
02 / Evidence
Benchmark results
14 distinct evaluations grouped by topic. Explore cross-model charts or switch to the full results list.
03 / Operation
Speed, tokens & price
Prices are per million tokens; latency and task costs describe the captured evaluation rather than a universal provider guarantee.
Token pricing
- Input / 1M
- $0.710
- Output / 1M
- $0.720
- Cache read / 1M
- $0.710
- Cache write / 1M
- —
- Cache-hit discount
- 50%
Response & usage
- Output speed
- 86.1 tok/s
- First chunk
- 2ms
- First answer
- 2ms
- Output / task
- —
- Reasoning / task
- —
Evaluation workload
- Canonical input
- —
- Canonical output
- —
- Answer / task
- —
- Index time / task
- —
- Run cost
- —
By prompt type
| Prompt | Output speed | First chunk | First answer | End to end |
|---|---|---|---|---|
| Medium | 87.7 tok/s | 1ms | 1ms | 6ms |
| Long | 86.1 tok/s | 2ms | 2ms | 7ms |
| Hundred K | 70.5 tok/s | 9ms | 9ms | 16ms |
| Medium Parallel | 84.1 tok/s | 1ms | 1ms | 7ms |
04 / Access
License & availability
Recorded license
LLAMA 3.3 COMMUNITY LICENSE AGREEMENT
The publisher-linked license text defines permitted use and redistribution. Review it before using the model.
- Weight access
- Open weights
- Classification
- Commercial License
- Reasoning
- No
Routing and providers
- OpenRouter ID
- meta-llama/llama-3.3-70b-instruct
- Tokenizer
- Llama3
- Instruct type
- llama3
- Captured endpoints
- 11
| Provider | Quantization | Context | tok/s |
|---|---|---|---|
| DeepInfra | fp8 | 131.1k | — |
| Novita | bf16 | 12.3k | — |
| AkashML | fp8 | 131.1k | — |
| Parasail | fp8 | 131.1k | — |
| Cloudflare | fp8 | 24k | — |
| SambaNova | unknown | 131.1k | — |
| Groq | unknown | 131.1k | — |
| CoreWeave | fp16 | 128k | — |
| unknown | 128k | — | |
| unknown | 128k | — | |
| Together | unknown | 131.1k | — |
05 / Audit
Underlying fields
Inspect the captured model and routing records. Artificial Analysis destination links are omitted.