Granite 4.1 8B
Apr 2026
Granite 4.1 8BCompare inon
Granite 4.1 8B measured on 16 of 104 Chat & Knowledge benchmarks16
BenchmarkByScoreRank
Overall1
Knowledge11
Science2
Documents1
Legacy · saturated or no longer reported1
No benchmark matches
—per task?
self-host onlyprice
—output speed?
131Kcontext
Alternatives?
ModelScoreCostPer taskWhat changes
BetterGPT-6 Lunalow×5.3—$0.005Function calling only with none · Chat Completions+3IOpenK2 Horizon 7Bthinking×4.8—$0.30 / $1.20Memory (4-bit): 4.4 GB → 4.5 GBBetterGranite 4.2 3BthinkingOpen+74%—$0.006Context: 131K → 128KBetterGPT-5.6 Lunalow×5.2—$0.01BetterGPT OSS 20BhighOpen×2.9—$0.011Memory (4-bit): 4.4 GB → 13 GBModelScoreCostPer taskWhat changes
Nothing here yet.
ModelScoreCostPer taskWhat changes
GPT-6 Lunalow×5.3—$0.005Function calling only with none · Chat Completions+3Granite 4.2 3BthinkingOpen+74%—$0.006Context: 131K → 128KGPT-5.6 Lunalow×5.2—$0.01GPT OSS 20BhighOpen×2.9—$0.011Memory (4-bit): 4.4 GB → 13 GB
Mistral Small 4thinkingOpen×2.6—$0.015Memory (4-bit): 4.4 GB → 72 GBModelScoreCostPer taskWhat changes
IK2 Horizon 7BthinkingOpen×4.8—$0.30 / $1.20Memory (4-bit): 4.4 GB → 4.5 GB
Gemma 4 12BthinkingOpen×4.1—$0.10 / $0.30Memory (4-bit): 4.4 GB → 6 GB
Qwen 3.5 9BthinkingOpen×3.9—$0.21Memory (4-bit): 4.4 GB → 4.8 GBIK2 Horizon 3.7BthinkingOpen×3.6—$0.30 / $1.20SApriel v1.5 15B ThinkerOpen×3.2——Context: 131K → 128K+1ModelScoreCostPer taskWhat changes
Granite 4.2 30BthinkingOpen×2.9—$0.072Context: 131K → 128K+1Granite 4.2 8BthinkingOpen×2.5—$0.024Context: 131K → 128KGranite 4.2 3BthinkingOpen+74%—$0.006Context: 131K → 128KGranite 4.0 H 350MOpen+70%——Context: 131K → 33KGranite 4.0 350Mnon-thinkingOpen+45%——Context: 131K → 33KGranite 4.0 H 1BOpen+32%——Context: 131K → 128KGranite 4.0 Micronon-thinkingOpen+32%—$0.017 / $0.11Granite 4.0 1Bnon-thinkingOpen+27%——Context: 131K → 128KAll IBM models →Run it yourself
estimateFits on✓ 8 GB (phone, laptop) · compressed✓ 16 GB · 8-bit✓ 24 GB GPU · full quality✓ 48 GB GPU · full quality✓ 80 GB GPU · full quality
LicenceApache-2.0free for commercial use
History
2 versionsBenchmarks
in Chat & KnowledgeAA Intelligence Index 4.3.2≈6.6#285
AA Omniscience Index-64.1#280
AA Omniscience · Accuracy12.3#265
Humanity's Last Exam3.8#316
GPQA Diamond43.3#272
AA-LCR (long context) 1.113.3#254
IFBench (instructions)38.6#166
CritPt (physics)0#180
FAQ
How much does Granite 4.1 8B cost?›
There is no per-unit API price: self-host only.
Is Granite 4.1 8B free to use?›
There is no free API. The weights are open, so you can run it yourself.
How good is Granite 4.1 8B, and what is it best at?›
#313 of 334 in Chat & Knowledge, #95 of 97 in Coding, #155 of 242 in Agents & Tool Use, #152 of 195 in Writing & Language.
Is Granite 4.1 8B better than Granite 4.0 H Small?›
On Humanity's Last Exam, Granite 4.1 8B scores 3.8 vs 3.8 — about the same.
What is the context window of Granite 4.1 8B?›
131K tokens.
When was Granite 4.1 8B released? Is there a newer version?›
Released 2026-04-29. It has been superseded by Granite 4.2 8B.
Is Granite 4.1 8B open source?›
Yes, the weights are open (Apache-2.0 licence).
Can I run Granite 4.1 8B locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Granite 4.1 8B read images?›
No, it accepts text only.
What is Granite 4.1 8B?›
Granite 4.1 8B is an open-weights language model from IBM, released Apr 2026.
What are the best alternatives to Granite 4.1 8B?›
GPT-6 Luna (better), K2 Horizon 7B (open weights).