All promotions
GuidesMCP

Granite 4.0 Micro

IBMReleased Oct 2, 2025granite-4.0-h-micro

Reasoning
no
Knowledge cutoff
April 2024
Context
128k tokens
API features
JSON mode, seed, logprobs

Read Oct 23, 06:00 UTC from the hosts' published offers and OpenRouter, spec sheet from IBM · cutoff: ibm.com

Input, per M$0.017at Cloudflare
Output, per M$0.11at Cloudflare
Speed35 tok/sat Cloudflare, on OpenRouter
Providers1Cloudflare only
ProviderInputOutputCache ↑tok/sFirst tokenUptime 24hChecked
Cloudflare$0.017$0.11—350.61 s100%Oct 23

1 provider · prices per million tokens, read Oct 23, 06:00 UTC · speed: OpenRouter, median of the last 30 min, read Oct 23, 06:00 UTC · uptime: OpenRouter, last 24 h

ProviderWeightsContextMax outputToolsStructured outputsReasoning
Cloudflare—131K—nonono

As served by each host and declared to OpenRouter; not verified by us · weights: the precision the host declares

Granite 4.0 Micro spec sheet, field by field, with sources

Granite 4.0 Micro spec sheetmaker-declared

What the maker declares about the model itself — read from its own weights repository and documentation, not from any host. What each endpoint actually serves is in the pricing above: the two can differ, and both are true.

Context window
to verify
Parameters
to verify
Active experts
to verify
Weight precision
to verify
License
to verifyScope: the licence on the published weights, and nothing else. It is not the maker's acceptable-use policy, which is a separate document, and it is not the contract you sign with whichever host you call — hosts set their own terms, which this index does not read.
Knowledge cutoff
April 2024
Modalities
to verify
Max output (official)
to verify
Release date
to verify≈ October 2, 2025 third-party catalogue — not the maker's figure
Maker lifecycle status
to verify

www.ibm.com · maker's documentation · 1 of 10 fields published by the maker

ProviderChatbot 1,000 in / 500 outRAG / long context 8,000 in / 500 outBatch summarize 4,000 in / 1,000 out
Cloudflare$0.07$0.19$0.18

What each host bills for 1,000 requests of a representative job

Your own workload

    Other IBM models · all its models, hosts and history ›