Granite 3.3 8B Instruct
Context
8K
Max output
8K
Serving providers
2
Cheapest input
$0.03 /1M
On-demand listed rates, OpenRouter routed quotes and Vercel reseller quotes. Enterprise commitments (committed-use discounts, provisioned throughput, negotiated rates) can be materially lower.
Benchmarks
independent evaluations · source: Artificial AnalysisHeadline indices
Individual benchmarks · click a row for effort variants
| Benchmark | Best score | Effort | Cost / task | |
|---|---|---|---|---|
| CritPt | 0% | default | — | |
| EvalComputeProxyFrozen Aug 18, 2026 | 1.9 | default | — | |
| GPQA Diamond | 34% | default | — | |
| Humanity's Last Exam | 4% | default | — | |
| IFBench | 22% | default | — | |
| LiveCodeBench | 13% | default | — | |
| Long-Context ReasoningArtificial Analysis did not report this score in its latest run on Oct 14, 2026 | 3% | — | — | |
| MMLU-ProFrozen Aug 18, 2026 | 47% | default | — | |
| Omniscience | -77.3 | default | — | |
| Omniscience Accuracy | 0.1 | default | — | |
| Omniscience Non Hallucination | 0.0 | default | — | |
| SciCodeArtificial Analysis did not report this score in its latest run on Oct 14, 2026 | 10% | — | — | |
| TerminalBench Hard | 0% | default | — | |
| τ²-bench | 11% | default | — |
Artificial Analysis last published scores for this benchmark on Aug 18, 2026. Scores here are frozen at that date, and models released since are not on it.
Pricing by serving provider
standard tier · per 1M tokens · click a provider to expandInput pricing runs from $0.03 to $0.20 per 1M tokens across 2 providers, so the dearest route costs 567% more than the cheapest for the same model.
| Serving provider | Input /1M | Output /1M | Endpoints | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ReplicateCheapest | $0.03 | $0.25 | 1 | ||||||||||
Standardreplicate/ibm-granite/granite-3.3-8b-instruct
| |||||||||||||
| Watsonx | $0.20 | $0.20 | 1 | ||||||||||
Standardwatsonx/ibm/granite-3-3-8b-instruct
| |||||||||||||
Price history
listed seller prices, input and output per 1M tokens Input Output
Input down 100% since first tracked
Cost calculator
estimate your monthly spend on this model$26
estimated / month
Model IDs
copy the exact identifier for your platformreplicate/ibm-granite/granite-3.3-8b-instructwatsonx/ibm/granite-3-3-8b-instruct
Frequently asked questions
Granite 3.3 8B Instruct pricing, context and availabilityHow much does Granite 3.3 8B Instruct cost?
Granite 3.3 8B Instruct costs $0.03 per 1M input tokens from Replicate, and $0.20 per 1M output tokens from Watsonx. Across 2 serving providers, input prices range from $0.03 to $0.20 per 1M tokens.
What is the context window of Granite 3.3 8B Instruct?
Granite 3.3 8B Instruct accepts up to 8K tokens of context and can return up to 8K output tokens.
Which providers serve Granite 3.3 8B Instruct?
Granite 3.3 8B Instruct is available from 2 serving providers, each with its own pricing and model ID. Replicate is currently the cheapest.
What can Granite 3.3 8B Instruct do?
Granite 3.3 8B Instruct supports tool use.
Other IBM models
compare pricing across the IBM lineupGranite 4.0 Micro$0.017/1M · 2 providersGranite 4.1 8B$0.05/1M · 2 providersGranite 4.2 8B$0.06/1M · 2 providersGranite 4 H Small$0.0636/1M · 1 providerGranite Guardian 3.2 2B$0.10/1M · 1 providerGranite Vision 3.2 2B$0.10/1M · 1 providerGranite 3 8B Instruct$0.20/1M · 1 providerGranite Guardian 3.3 8B$0.20/1M · 1 provider
Listed prices from LiteLLM, OpenRouter and Vercel · last checked Oct 14, 2026
Every weekday
AI moves fast. Here's your debrief.
News, analysis, tools, and more.
For people who build with AI