Granite 4.2 30B
- Input
- $0.16/M tok
- Output
- $0.65/M tok
- Blended
- $0.282/M tok · 3:1
- Speed
- 77tok/s
- Context
- 131Ktokens
- Cost per task
- ~$0.27/taskest.
Long-tail import from Artificial Analysis (2026-09-24); AA variant: Granite 4.2 30B. 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
- BF16 weights
- Throughput
- ~4,066 tok/s
- many requests batched
- Price
- $0.23–2.03 /M tok
- busy → light use
- On your own machine
- RTX 5090 32GB (int4, ~86 tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- BF16 weights (29.3B params) + 25% KV-cache headroom ≈ 73 GB
- 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
- ~4,066 output tok/s aggregate at batch 116 (bandwidth-bound)
- Blended 3:1 input:output; prefill ~18,142 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.
Coding
SWE-bench VerifiedEstimated
~#17 of 21 (estimated)~68.4%
- #1Claude Opus 596.0%
- #2Claude Sonnet 585.2%
- #16Nemotron 3 Ultra70.7%
- ~#17Granite 4.2 30BEstimated68.4%
- #17Intern-S2-Preview (35B-A3B)64.0%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#23 of 24 (estimated)~47.0%
- #1Claude Fable 5.293.0%
- #2Claude Opus 5.589.9%
- #22Muse Glimmer51.2%
- ~#23Granite 4.2 30BEstimated47.0%
- #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#41 of 46 (estimated)~41.8%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #40Qwen3.8 27B42.2%
- ~#41Granite 4.2 30BEstimated41.8%
- #41Laguna S 2.140.4%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Verified
#63 of 1010.0%
- #1Kimi K3.1 Pro67.2%
- #2Claude Sonnet 5.563.6%
- #63gpt-oss-20b0.0%
- #63Granite 4.2 30B0.0%
- #63Granite 4.2 3B0.0%
artificialanalysis.ai · 2026-10-10
LiveCodeBenchEstimated
~#17 of 28 (estimated)~55.8%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #16Claude Haiku 4.561.5%
- ~#17Granite 4.2 30BEstimated55.8%
- #17GLM-5.3 NVFP453.5%
No published score — estimated from related benchmarks
SciCodeVerified
#90 of 11837.8%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #88Mercury 2.538.5%
- #90Granite 4.2 30B37.8%
- #91G9v3-39A5B36.8%
artificialanalysis.ai · 2026-09-24
AA Coding Agent IndexEstimated
~#25 of 26 (estimated)~28.1
- #1Claude Opus 5.561.7
- #2Claude Opus 559.7
- #24GPT-6 Luna31.1
- ~#25Granite 4.2 30BEstimated28.1
- #25Muse Spark Open26.3
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondVerified
#127 of 19264.4%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #126Solar Open 100B65.7%
- #127Granite 4.2 30B64.4%
- #128LongCat Flash Lite63.6%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#104 of 19811.2%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #103NVIDIA Nemotron 3 Nano 30B A3B11.4%
- #104Granite 4.2 30B11.2%
- #105North Mini Code11.1%
artificialanalysis.ai · 2026-09-24
AIME (latest)Estimated
~#23 of 34 (estimated)~57.5%
- #1GLM-5.3 NVFP499.5%
- #2Beam 1.0 501B A23B97.7%
- #22Phi-4-reasoning-plus78.0%
- ~#23Granite 4.2 30BEstimated57.5%
- #23Mistral Large 338.0%
No published score — estimated from related benchmarks
Agents
τ²-bench (Telecom)Estimated
~#54 of 120 (estimated)~44.7%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #53Nemotron 3 Nano Omni 30B A3B Reasoning45.3%
- ~#54Granite 4.2 30BEstimated44.7%
- #54Gemma 4 26B A4B43.6%
No published score — estimated from related benchmarks
OSWorld-VerifiedEstimated
~#12 of 15 (estimated)~60.9%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #11Muse Glimmer65.9%
- ~#12Granite 4.2 30BEstimated60.9%
- #12Nemotron 3.5 Super VL NVFP452.3%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~last of 5 (estimated)below measured range (<47.5%)
No published score — estimated from related benchmarks
BrowseCompEstimated
~#19 of 22 (estimated)~57.4%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #18Nemotron 3.5 Bolt 74B-A9B70.0%
- ~#19Granite 4.2 30BEstimated57.4%
- #19Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Verified
#64 of 91564 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #63gpt-oss-120b596
- #64Granite 4.2 30B564
- #65Gemma 4 26B A4B553
artificialanalysis.ai · 2026-09-24