Skip to content
Model Pareto

Granite 4.2 3B

IBM GraniteOpen weightsartificialanalysis.ai ↗
Input
$0.03/M tok
Output
$0.12/M tok
Blended
$0.052/M tok · 3:1
Speed
220tok/s
Context
131Ktokens
Cost per task
~$0.05/taskest.

Long-tail import from Artificial Analysis (2026-09-24); AA variant: Granite 4.2 3B. 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
~16,546 tok/s
many requests batched
Price
$0.045–0.40 /M tok
busy → light use
On your own machine
RTX 5090 32GB (int4, 300+ tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • BF16 weights (3.66B params) + 25% KV-cache headroom ≈ 9 GB
  • 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~16,546 output tok/s aggregate at batch 256 (bandwidth-bound)
  • Blended 3:1 input:output; prefill ~145,236 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

~#23 of 24 (estimated)~45.4%
  1. #1Claude Fable 5.293.0%
  2. #2Claude Opus 5.589.9%
  3. #22Muse Glimmer51.2%
  4. ~#23Granite 4.2 3BEstimated45.4%
  5. #23Nex-N2.5-Mini43.8%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#43 of 46 (estimated)~32.7%
  1. #1Gemini 4 Argon78.8%
  2. #2DeepSeek V4.1 Flash 101777.0%
  3. #42Nex-N2.5-Mini36.1%
  4. ~#43Granite 4.2 3BEstimated32.7%
  5. #43Hy328.0%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#22 of 28 (estimated)~44.7%
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #2Qwen3.8 Flash-Next91.9%
  3. #21Devstral 244.8%
  4. ~#22Granite 4.2 3BEstimated44.7%
  5. #22Llama 4 Maverick39.7%
No published score — estimated from related benchmarks
~#25 of 26 (estimated)~27.9
  1. #1Claude Opus 5.561.7
  2. #2Claude Opus 559.7
  3. #24GPT-6 Luna31.1
  4. ~#25Granite 4.2 3BEstimated27.9
  5. #25Muse Spark Open26.3
No published score — estimated from related benchmarks

Hard reasoning

Agents