Llama 3.3 Instruct 70B
- Input
- $0.71/M tok
- Output
- $0.72/M tok
- Blended
- $0.712/M tok · 3:1
- Speed
- 85tok/s
- Context
- 128Ktokens
Long-tail import from Artificial Analysis (2026-09-24); AA variant: Llama 3.3 Instruct 70B. Price = AA median host price.
Rankings
Self-hostingEstimated
How we estimate →What it would cost to run these open weights yourself on rented GPUs, compared with the API price above.
- Hardware
- 1× B200 180GB
- FP8 weights
- Throughput
- ~2,313 tok/s
- many requests batched
- Price
- $0.35–3.12 /M tok
- busy → light use
- On your own machine
- Mac Studio M5 Max 128GB (int4, ~12 tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- FP8 weights (70.6B params) + 25% KV-cache headroom ≈ 88 GB
- 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
- ~2,313 output tok/s aggregate at batch 66 (bandwidth-bound)
- Blended 3:1 input:output; prefill ~15,058 tok/s
- Low = 75% utilization at the low GPU price; high = 20% at the high price
Benchmark scores
~ italic, dashed = no published score yet, estimated from related benchmarks.
Writing & chat
IFBenchVerified
#76 of 13947.1%
- #1MiniMax-M382.9%
- #2Qwen3.8 2.4T A95B82.8%
- #75Mistral Small 448.2%
- #76Llama 3.3 Instruct 70B47.1%
- #77Cogito v2.146.3%
artificialanalysis.ai · 2026-09-24
MMMLUEstimated
~last of 5 (estimated)below measured range (<81.3%)
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Estimated
~#49 of 57 (estimated)~1310 Elo
- #1Gemini 4 ArgonFew votes1512
- #2Claude Opus 5.51497
- #48Llama 4 Maverick1311
- ~#49Llama 3.3 Instruct 70BEstimated1310
- #49Llama 3.1 Nemotron Ultra 253B v1Few votes1306
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#49 of 57 (estimated)~1334 Elo
- #1Gemini 4 Argon1525
- #2Claude Opus 5.51507
- #48Llama 3.1 Nemotron Ultra 253B v1Few votes1348
- ~#49Llama 3.3 Instruct 70BEstimated1334
- #49Olmo 3.1 32B Instruct1330
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#22 of 26 (estimated)~1486 Elo
- #1GPT-6 Astra2173
- #2Claude Fable 5.12162
- #21Gemini 3.1 Pro1491
- ~#22Llama 3.3 Instruct 70BEstimated1486
- #22Mistral Large 31412
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondVerified
#155 of 19249.8%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #153NVIDIA Nemotron 3 Nano 4B51.3%
- #155Llama 3.3 Instruct 70B49.8%
- #156Falcon-H1-34B-Instruct49.7%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#192 of 1983.6%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #192Hermes 4 - Llama-3.1 70B3.6%
- #192Llama 3.3 Instruct 70B3.6%
- #196Devstral Small 23.5%
artificialanalysis.ai · 2026-09-24
AA Intelligence IndexEstimated
~#100 of 103 (estimated)~4.9
- #1Gemini 4 Carbon59.4
- #2Claude Opus 5.557.6
- #99Ministral 3 8B5.5
- ~#100Llama 3.3 Instruct 70BEstimated4.9
- #100Ministral 3 3B4.8
No published score — estimated from related benchmarks
Agents
τ²-bench (Telecom)Verified
#66 of 11926.6%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #66INTELLECT-326.6%
- #66Llama 3.3 Instruct 70B26.6%
- #66Ministral 3 8B26.6%
artificialanalysis.ai · 2026-09-24
OSWorld-VerifiedEstimated
~#13 of 15 (estimated)~52.0%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #12Nemotron 3.5 Super VL NVFP452.3%
- ~#13Llama 3.3 Instruct 70BEstimated52.0%
- #13Nemotron 3.5 Super VL51.1%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~#3 of 5 (estimated)~53.8%
No published score — estimated from related benchmarks
BrowseCompEstimated
~#21 of 22 (estimated)~38.8%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #20Nemotron 3 Ultra44.4%
- ~#21Llama 3.3 Instruct 70BEstimated38.8%
- #21Mistral Small 421.3%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Estimated
~#60 of 92 (estimated)~698 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #59Claude Haiku 4.5719
- ~#60Llama 3.3 Instruct 70BEstimated698
- #60MiniCPM5-2B694
No published score — estimated from related benchmarks
AutomationBench-AAEstimated
~#79 of 85 (estimated)~0.2%
- #1Claude Fable 5.276.4%
- #2Muse Wonder Full71.3%
- #77Granite 4.2 3B0.4%
- ~#79Llama 3.3 Instruct 70BEstimated0.2%
- #79gpt-oss-120b0.2%
No published score — estimated from related benchmarks
τ-Bench Banking (AA)Estimated
~#91 of 91 (estimated)~3.2%
No published score — estimated from related benchmarks