Granite 4.0 Micro
IBM · released Sep 16, 2025 · ibm-granite/granite-4.0-h-micro
- Type
- Open weightsApache License 2.0
- Params
- 3.2B
- Context
- 131K
about 98K words of context
Our take
Written Aug 3, 2026Granite 4.0 Micro is a tiny downloadable text model from IBM with a permissive Apache licence. It is built for cost-sensitive deployment where a 131,000-token request limit is enough and no images or audio are needed.
Pick this for minimal-cost text tasks where inference price matters more than measured quality, or for edge deployment that needs a permissive licence. Use it when you know a 131,000-token limit fits your longest prompt and you do not need multimodal input. Skip it if you need benchmark-backed quality claims, image or audio support, or if you expect to self-host and need evidence of efficiency tricks to offset the small size.
The case for it
- Extremely low inference cost, with identical pricing across both tracked providers.
- Apache 2.0 licence allows commercial use, modification and redistribution.
- One hosted route lists throughput at 29 tokens per second.
The case against it
- No measured quality benchmarks in our data — no Elo, MMLU or other scores held.
- Very small at 3.2 billion parameters, with no disclosed active-parameter figure or sparse-architecture claim to offset the size.
- Text-to-text only; no image, video or audio support.
How good is it?
We hold no score for this model.
We look for every model we track on every board we watch, and none of them has turned up Granite 4.0 Micro — so there is no intelligence, coding, agentic or writing score to show you, not a low one, none.
Can you run it yourself?
- Fits in memory
- weights load entirely on the card
- Spills to system RAM
- some weights offload; much slower
- Too large
- will not load even with offload
- est
- size is calculated; the verdict could change by 10%
Comfortable fit
GeForce RTX 4090 · 24 GB
Room to spare. 19.4 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 27.4 GB spare means a 10% error in the size would not change the answer.
Apple M2 (8-core GPU, 8GB unified) · 8 GB
Room to spare. 2.6 GB spare means a 10% error in the size would not change the answer.
Memory use by level
Against a 24 GB card.
All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.
Check against your own machine → · All 71 devices, with every size →
Or rent it from someone else
Cheapest of 2 live listings. Picked at the widest standard context we hold, within one quantisation slice, so the numbers beside it are a price one host actually charges.
- per 1M tokens
- $0.017 in / $0.11 out
- Context served
- 131K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouter | $0.017 / $0.11 | 131K | not measured | Unknown | Unknown | Unknown |
| Cloudflare Workers AI | $0.017 / $0.11 | 131K | 30 tok/s | No | Yesunknown period | Unknown |
Cloudflare Workers AIAcross the 2 listings we hold: 1 say they do not train on prompts, 0 say they do and 1 do not say. 0 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouter | ✗ | ✓ | ✗ |
| Cloudflare Workers AI | ✗ | ✓ | ✗ |
Tool calling: 0 of 2 listings say yes, 2 say no. JSON output: 2 of 2 listings say yes. Strict schema: 0 of 2 listings say yes, 2 say no.
When we formed this view
Dates behind this page
Prices last checked 7h ago
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- No board we watch has turned up a score, so we hold no quality figures at all.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 2 listings do not say whether they train on prompts.
- We hold no cached-input rate for any of its listings.
Licence and identifiers
What the licence allowsApache License 2.0, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
Apache License 2.0
Fully permissive: commercial use, redistribution, and derivatives allowed. Requires attribution and a copy of the license. Includes an express patent grant.
Identifiers
- Hugging Face
- ibm-granite/granite-4.0-h-micro
- Architecture
- Mixture of experts
- Modality record
- text->text
- Catalogue slug
- ibm-granite-granite-4-0-micro