Llama 3.3 70B Instruct llama-3.3-70b-instruct
meta-llama · llama
#94 overall
- params
- 70.6B
- arch
- dense
- context
- 128k
- license
- llama3.3 open weights gated
- released
- Dec 2024
- downloads/30d
- 300.2k
# architecture
signal path of one layer, generated from the registry's structured fields — dimension lines quote the model's real numbers; an MoE trace forks at the router, a dense trace runs straight through. Geometry fields come from the repo's config.json.
# why ranked
Overall: #94 — score 13.5 ● low — a single signal; treat with caution
| signal | weight | input (0–100) |
|---|---|---|
| aa_intelligence | 0.70 | 14.6 |
| bench_composite | 0.30 | 11.0 |
benchmark panel evidence:
| complete panel | score (0–100) |
|---|---|
| aa-coding-v1 | 21.9 |
| aa-math-v1 | 0.0 |
Coding: provisional — score 11.8 ● low — a single signal; treat with caution
provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.
| signal | weight | input (0–100) |
|---|---|---|
| aa_coding | 0.30 | 11.8 |
| aider_polyglot | 0.30 | — |
| swe_bench_verified | 0.40 | — |
missing signals are dropped and the remaining weights renormalized — never imputed.
# trend
methodology changed during this history window; score movement across that boundary is not model movement. See methodology v6.
# benchmarks
| benchmark | score | source | date |
|---|---|---|---|
| AA Coding Index | 11.9 | Artificial Analysis | — |
| AA Intelligence Index | 7.7 | Artificial Analysis | — |
| AA Math Index | 7.7 | Artificial Analysis | — |
| Bfcl-v2 | 0.8 / 1 | LLM Stats | — |
# where to run
| provider | quant | ctx | $/M in | $/M out | $/M cache | price src | tps | uptime | ✓ |
|---|---|---|---|---|---|---|---|---|---|
| Featherless | — | — | — | — | via hfrouter | — | — | ||
| DeepInfra | fp8 | 128k | $0.10 | $0.32 | — | via openrouter | — | 99.6% | |
| Nebius | 125k | $0.13 | $0.40 | $0.13 | via requesty | — | — | ||
| Crusoe | 128k | $0.20 | $0.20 | — | via litellm | — | — | ||
| Nscale | — | $0.20 | $0.20 | — | via litellm | — | — | ||
| Novita | bf16 | 12k | $0.14 | $0.40 | — | novita | — | 99.9% | ✓ |
| Inferenceapis | 128k | $0.20 | $0.50 | — | via litellm | — | — | ||
| AkashML | fp8 | 128k | $0.20 | $0.52 | $0.10 | via openrouter | — | 100.0% | |
| Parasail | fp8 | 128k | $0.22 | $0.50 | $0.11 | via openrouter | — | 99.8% | |
| Hyperbolic | 128k | $0.40 | $0.40 | — | hyperbolic | — | — | ✓ | |
| SambaNova | unknown | 128k | $0.45 | $0.90 | — | via openrouter | — | 100.0% | |
| Groq | unknown | 128k | $0.59 | $0.79 | $0.29 | via openrouter | — | 100.0% | |
| Azure | 125k | $0.71 | $0.71 | — | via modelsdev | — | — | ||
| CoreWeave | fp16 | 125k | $0.71 | $0.71 | $0.71 | via openrouter | — | 100.0% | |
| Wandb | 125k | $0.71 | $0.71 | — | via litellm | — | — | ||
| unknown | 125k | $0.72 | $0.72 | — | via openrouter | — | — | ||
| unknown | 125k | $0.72 | $0.72 | — | via openrouter | — | — | ||
| OVHcloud | 128k | $0.74 | $0.74 | — | via hfrouter | 19 | — | ||
| Cloudflare | fp8 | 24k | $0.29 | $2.25 | — | via openrouter | — | 100.0% | |
| Scaleway | — | $1.03 | $1.03 | — | via hfrouter | 91 | — | ||
| Venice | fp8 | 125k | $0.70 | $2.80 | — | venice | — | — | ✓ |
| Phala | 128k | $2.00 | $2.00 | — | phala | — | — | ✓ |
sorted by blended price ((3·input + output) / 4 per 1M) · ✓ = the provider's own catalog confirms the offer · "via …" prices are what the aggregator routing the offer charges, not the provider's own list price
source aliases
- aa
llama-3-3-70b-instruct,llama-3-3-instruct-70b- arena
Llama-3.3-70B-Instruct- openrouter
meta-llama/llama-3.3-70b-instruct