IBMGranite 3.3 8B
As of 29 Oct 2026, Granite 3.3 8B by IBM is not yet ranked on the BenchLeader Index (3 independent results so far; the index needs at least five across two categories). It costs $0.085 per 1M tokens blended. Released 16 Apr 2025; 3 independent results so far.
Key facts (as of 29 Oct 2026)
- Granite 3.3 8B costs $0.085 per 1M tokens blended (list price, 29 Oct 2026).
- Granite 3.3 8B was released on 16 Apr 2025 (IBM).
In detail. Granite 3.3 8B is an IBM open-weights model, released 16 Apr 2025. It is not yet ranked: 3 independent results so far, and the index needs at least five across two categories. It has been measured at 2 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 24 Nov 2025.stale: no new result in six months
- Blended price
- –
- Output speed
- –
- First answer
- –
- Context
- –
- Full answer
- –
- Cost per run
- –
- Released
- 16 Apr 2025
- Maths32
- Instruction following36
Versions
IBM has shipped 6 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Granite 4.2 8B | 25 Aug 2026 | 43.9 | #459 |
| Granite 4.1 8B | 29 Apr 2026 | 42.3 | #478 |
| Granite 3.3 8Bthis page | 16 Apr 2025 | – | – |
| Granite 3.1 8B | 18 Dec 2024 | 41.7 | #489 |
| Granite 3.0 8B | – | – | – |
| Granite 3.2 8B | – | – | – |
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | – | – | – | – | – | Agents & tools 34 · Instruction following 27 |
| not statedbest | – | – | – | – | – | Instruction following 36 · Maths 32 |
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | no reasoning | not stated | Source |
|---|---|---|---|
| GPQA Diamond (AA)not in index | 33.8%#505 | – | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 4.2%#542 | – | Artificial Analysis |
| GPQA Diamond (HELM)not in indexeffort not stated | – | 32.5%#58 | HELM Capabilities |
| CritPtnot in index | 0.0%#328 | – | Artificial Analysis |
Coding
| Benchmark | no reasoning | Source |
|---|---|---|
| LiveCodeBench (AA)not in index | 12.7%#266 | Artificial Analysis |
Agents & tools
| Benchmark | no reasoning | Source |
|---|---|---|
| Terminal-Bench Hard | 0.0%#364 | Artificial Analysis |
| τ²-Bench Telecom (AA)not in index | 10.5%#384 | Artificial Analysis |
Maths
| Benchmark | not stated | Source |
|---|---|---|
| Omni-MATH (HELM) | 17.6%#58 | HELM Capabilities |
Knowledge
| Benchmark | not stated | Source |
|---|---|---|
| MMLU-Pro (HELM)not in index | 34.3%#62 | HELM Capabilities |
Instruction following
| Benchmark | no reasoning | not stated | Source |
|---|---|---|---|
| IFBench | 22.4%#406 | – | Artificial Analysis |
| IFEval (HELM)effort not stated | – | 72.9%#62 | HELM Capabilities |
| WildBench (HELM)effort not stated | – | 74.1%#53 | HELM Capabilities |
Composite
| Benchmark | no reasoning | not stated | Source |
|---|---|---|---|
| AA Intelligence Index v4.3.2not in index | 4.9#624 | – | Artificial Analysis |
| AA Progress Index v1not in index | 130.9#583 | – | Artificial Analysis |
| HELM Capabilities meannot in indexeffort not stated | – | 46.3%#61 | HELM Capabilities |
More from IBM
See also
Questions people ask
- Why does Granite 3.3 8B have no BenchLeader Index score?
- The index needs at least five independent results across at least two categories; as of 29 Oct 2026 Granite 3.3 8B has 3 that count. Launch-post scores are never used, so it is ranked once independent publishers have measured it.
- When was Granite 3.3 8B released?
- IBM released Granite 3.3 8B on 16 Apr 2025. BenchLeader has results for it at 2 reasoning-effort settings.
Data as of 29 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Granite 3.3 8B: benchmarks, pricing, speed and rank”, data as of 29 Oct 2026, https://www.benchleader.com/models/granite-3-3-8b.