Skip to content

Model Library

IBMReleased April 16, 2025

Granite 3.3 8B

IBM · Released April 16, 2025

View public research Browse the library

Intelligence score

2

At a glance

Intelligence
2
Speed
385.4tokens/sec
Latency
28.20stime to first token
Input price
$0.03per 1M tokens
Output price
$0.25per 1M tokens
Context window
130Ktokens
Parameters
8.17billions

Summary

Comparison summary

Granite 3.3 8B (Non-reasoning) scores 2 (estimated) on the Artificial Analysis Intelligence Index, placing it at the lower end among other open weight non-reasoning models of similar size (median: 6). Granite 3.3 8B (Non-reasoning) generates output at 385.4 tokens per second (based on the median across providers serving the model), which is well above average compared to other open weight non-reasoning models of similar size (median: 101.8 t/s). Granite 3.3 8B (Non-reasoning) costs $0.03 per 1M input tokens (very competitive, median: $0.15) and $0.25 per 1M output tokens (better than average, median: $0.32), based on the median across providers serving the model.

Specifications

Technical specifications

Model nameGranite 3.3 8B
ProviderIBM
Release dateApril 16, 2025
Intelligence score2
Speed385.4 tokens/sec
Latency (TTFT)28.20s
Input price$0.03 / 1M tokens
Output price$0.25 / 1M tokens
Context window130,000 tokens
Parameters8.17B
ReasoningNo
Input modalitiestext
Output modalities—
Open weightsYes
LicenseApache 2
API availability—
SourceModel Intelligence legacy profile: granite-3-3-8b-instruct

Pricing

Token prices

Price per 1M tokens

Granite 3.3 8B, as listed in the catalog record

Input $0.03 and output $0.25 per 1M tokens. Missing prices are hatched and marked "Not published".
  1. Input$0.03
  2. Output$0.25

Model Intelligence legacy profile: granite-3-3-8b-instruct

FAQ · 19

Frequently asked questions

Granite 3.3 8B (Non-reasoning) was released on April 16, 2025.

Granite 3.3 8B (Non-reasoning) was created by IBM.

Granite 3.3 8B (Non-reasoning) scores 2 (estimated) on the Artificial Analysis Intelligence Index, placing it at the lower end among other open weight non-reasoning models of similar size (median: 6).

Granite 3.3 8B (Non-reasoning) generates output at 385.4 tokens per second (based on the median across providers serving the model), which is well above average compared to other open weight non-reasoning models of similar size (median: 101.8 t/s).

Granite 3.3 8B (Non-reasoning) has a time to first token (TTFT) of 28.20s (based on the median across providers serving the model), which is at the higher end compared to other open weight non-reasoning models of similar size (median: 1.54s).

Granite 3.3 8B (Non-reasoning) costs $0.03 per 1M input tokens (very competitive, median: $0.15) and $0.25 per 1M output tokens (better than average, median: $0.32), based on the median across providers serving the model.

Granite 3.3 8B (Non-reasoning) costs $0.03 per 1M input tokens and $0.25 per 1M output tokens (based on the median across providers serving the model). For a blended rate (7:2:1 cache hit/input/output ratio), this is $0.05 per 1M tokens. Pricing may vary by provider.

No, Granite 3.3 8B (Non-reasoning) is not a reasoning model. It provides direct responses without extended chain-of-thought reasoning.

Granite 3.3 8B (Non-reasoning) supports text input.

Granite 3.3 8B (Non-reasoning) supports text output.

No, Granite 3.3 8B (Non-reasoning) does not support image input. It can only process text.

No, Granite 3.3 8B (Non-reasoning) is not multimodal. It only supports text input.

Granite 3.3 8B (Non-reasoning) has a context window of 130k tokens. This determines how much text and conversation history the model can process in a single request.

Yes, Granite 3.3 8B (Non-reasoning) is open weights. The model weights are publicly available and can be downloaded for self-hosting.

Granite 3.3 8B (Non-reasoning) has 8.17 billion parameters.

Granite 3.3 8B (Non-reasoning) is released under the Apache 2.0 license. This license allows commercial use.

Granite 3.3 8B (Non-reasoning) achieves a score of 2 on the Artificial Analysis Intelligence Index. This composite benchmark evaluates models across reasoning, knowledge, mathematics, and coding.

Yes, Granite 3.3 8B (Non-reasoning) is available via API through 1 provider.

Granite 3.3 8B (Non-reasoning) is available through 1 API provider.