Skip to main content
Back to Explore

Granite 4.0 H Small

IBM·Released 2025-09-22
Open Source
Compare

Related Models

IBM: Granite 4.1 8B2026-04-30Granite 4.1 30B2026-04-29Granite 4.1 3B2026-04-29Granite 4.0 H 1B2025-10-28Granite 4.0 1B2025-10-28Granite 4.0 350M2025-10-28Granite 4.0 H 350M2025-10-28IBM: Granite 4.0 Micro2025-10-20

Pricing

Input

$0.06

per 1M tokens

Output

$0.25

per 1M tokens

Blended

$0.11

per 1M tokens

Cheaper than 82% of models. Median price is $0.54/1M tokens.

Cost Calculator

Tokens per day1M
100K100M

Daily

$0.11

Monthly

$3.21

vs. Similar Models

Jamba 1.7 LargeQ:+0.1
$3.50+3171%
Jamba 1.5 LargeQ:-0.1
$3.50+3171%
Nous: Hermes 3 70B InstructQ:-0.1
$0.70+554%
Qwen3 8B (Non-reasoning)Q:-0.1
$0.18+73%

Performance

441

tokens/sec

Faster than 99% of models

8.72

seconds

Faster than 13% of models

8.72

seconds

Faster than 42% of models

Market Median

94 tok/s

367% faster

Median TTFT

1.11s

684% slower

Throughput/Dollar

4117

tok/s per $/1M

Speed Comparison

HyperNova 60B 2605
417 tok/s-5%
StepFun: Step 3.7 Flash
400 tok/s-9%
LFM2.5-VL-1.6B
508 tok/s+15%

Benchmarks

MMLU-Pro
62.4%
GPQA Diamond
41.6%
HLE
3.7%
LiveCodeBench
25.1%
SciCode
20.9%
TerminalBench Hard
2.3%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025
13.7%
IFBench
31.5%
Long Context Recall
9.0%
Tau2
17.3%
Market AverageTop Score

Open Source

Quick Compare

Similar Models

DeepHermes 3 - Mistral 24B Preview (Non-reasoning)

Nous Research

Q: 5.3N/A/1M

Jamba 1.7 Large

AI21 Labs

Q: 5.3$3.50/1M
Slower: 86%Pricier: 3171%

Nous: Hermes 3 70B Instruct

Nous Research

Q: 5.1$0.70/1M131K ctx
Slower: 93%Pricier: 554%

Qwen3 Omni 30B A3B Instruct

Alibaba

Q: 5.1$0.43/1M
Slower: 76%Pricier: 302%

DeepSeek-Coder-V2

DeepSeek

Q: 5.1N/A/1M

Jamba 1.5 Large

AI21 Labs

Q: 5.1$3.50/1M
Pricier: 3171%
Compare all 7 models