Granite 4.0 H Micro vs LFM2.5-24B-A2B
Granite 4.0 H Micro (2025) and LFM2.5-24B-A2B (2026) are frontier reasoning models from IBM Research and Liquid AI. Granite 4.0 H Micro ships a 131K-token context window, while LFM2.5-24B-A2B ships a 128K-token context window. On Google-Proof Q&A, LFM2.5-24B-A2B leads by 42.3 pts. On pricing, Granite 4.0 H Micro costs $0.02/1M input tokens versus $0.05/1M for the alternative. This comparison covers specs, pricing, API access, capabilities, benchmarks, input and output token costs, and production fit for coding and agent workloads.
Granite 4.0 H Micro is ~194% cheaper at $0.02/1M; pay for LFM2.5-24B-A2B only for reasoning depth.
Decision scorecard
Local evidence first| Signal | Granite 4.0 H Micro | LFM2.5-24B-A2B |
|---|---|---|
| Best for | tool-calling agents and provider-routed production | reasoning-heavy apps, tool-calling agents, and provider-routed production |
| Decision fit | Coding, RAG, and Agents | RAG, Agents, and Long context |
| Context window | 131K | 128K |
| Cheapest output | $0.11/1M tokens | $0.70/1M tokens |
| Provider routes | 2 tracked | 3 tracked |
| Shared benchmarks | 1 shared | Google-Proof Q&A leader |
Decision tradeoffs
- Granite 4.0 H Micro has the lower cheapest tracked output price at $0.11/1M tokens.
- Local decision data tags Granite 4.0 H Micro for Coding, RAG, and Agents.
- LFM2.5-24B-A2B holds a shared-benchmark lead on Google-Proof Q&A, ahead by 42.3 points.
- LFM2.5-24B-A2B has the larger context window for long prompts, retrieval packs, or transcript analysis.
- LFM2.5-24B-A2B has broader tracked provider coverage for fallback and route flexibility.
- LFM2.5-24B-A2B uniquely exposes Reasoning and Structured outputs in local model data.
- Local decision data tags LFM2.5-24B-A2B for RAG, Agents, and Long context.
Monthly cost at traffic
Estimate token spend from the cheapest tracked input and output route or tier on this page.
Granite 4.0 H Micro
$41.60
Cheapest tracked route/tier: Cloudflare Workers AI
LFM2.5-24B-A2B
$215
Cheapest tracked route/tier: Liquid AI
Estimated monthly gap: $173. Batch, cache, alternate speed tiers, and negotiated pricing are excluded from this local estimate.
Switch friction
- Provider overlap exists on OpenRouter; start route-level A/B tests there.
- LFM2.5-24B-A2B is $0.59/1M tokens higher on cheapest tracked output pricing, so quality gains need to justify the spend.
- LFM2.5-24B-A2B adds Reasoning and Structured outputs in local capability data.
- Provider overlap exists on OpenRouter; start route-level A/B tests there.
- Granite 4.0 H Micro is $0.59/1M tokens lower on cheapest tracked output pricing before cache, batch, or negotiated discounts.
- Check replacement coverage for Reasoning and Structured outputs before moving production traffic.
Specs
| Specification | ||
|---|---|---|
| Released | 2025-04-01 | 2026-10-13 |
| Context window | 131K | 128K |
| Parameters | — | 24B |
| Architecture | Decoder Only | Mixture of Experts |
| License | Apache 2.0OSI-approved | LFM Open License v1.0 |
| Openness | Open source | Open weights |
| Weights | Available | Available |
| Code | Unknown | Unknown |
| Commercial use | Permitted | Restricted |
| Knowledge cutoff | - | - |
Pricing and availability
| Pricing attribute | Granite 4.0 H Micro | LFM2.5-24B-A2B |
|---|---|---|
| Input price | $0.02/1M tokens | $0.05/1M tokens |
| Output price | $0.11/1M tokens | $0.70/1M tokens |
| Providers |
Capabilities
| Capability | Granite 4.0 H Micro | LFM2.5-24B-A2B |
|---|---|---|
| Vision | No | No |
| Multimodal | No | No |
| Reasoning | No | Yes |
| JSON / Tool use | Yes | Yes |
| Structured outputs | No | Yes |
| Code execution | No | No |
| IDE integration | No | No |
| Computer use | No | No |
| Parallel agents | No | No |
Benchmarks
| Benchmark | Granite 4.0 H Micro | LFM2.5-24B-A2B |
|---|---|---|
| Google-Proof Q&A | 32.1 | 74.4 |
Deep dive
On shared benchmark coverage, Google-Proof Q&A has Granite 4.0 H Micro at 32.1 and LFM2.5-24B-A2B at 74.4, with LFM2.5-24B-A2B ahead by 42.3 points. The largest visible gap is 42.3 points on Google-Proof Q&A, which matters most when that benchmark mirrors your workload. Treat isolated benchmark wins as directional, because provider routing, prompt style, and tool access can move real application results.
The capability footprint differs most on reasoning mode: LFM2.5-24B-A2B and structured outputs: LFM2.5-24B-A2B. Both models share JSON / Tool use, so the practical split is not just feature count. Use those differences to decide whether the page is about raw model quality, agentic coding support, multimodal ingestion, or predictable structured API behavior.
For cost, Granite 4.0 H Micro lists $0.02/1M input and $0.11/1M output tokens on the cheapest listed provider, while LFM2.5-24B-A2B lists $0.05/1M input and $0.70/1M output tokens on the cheapest listed provider. A 70/30 input-output blend puts Granite 4.0 H Micro lower by about $0.20 per million blended tokens. Availability is 2 providers versus 3, so concentration risk also matters.
Choose Granite 4.0 H Micro when provider fit and lower input-token cost are central to the workload. Choose LFM2.5-24B-A2B when reasoning depth, larger context windows, and broader provider choice are more important. For production, rerun your own prompts through the exact provider, region, and tool stack you plan to ship.
FAQ
Which has a larger context window, Granite 4.0 H Micro or LFM2.5-24B-A2B?
LFM2.5-24B-A2B supports 128K tokens, while Granite 4.0 H Micro supports 131K tokens. That gap matters most for long documents, large codebases, retrieval-heavy agents, and conversations where earlier context must remain visible.
Which is cheaper, Granite 4.0 H Micro or LFM2.5-24B-A2B?
Granite 4.0 H Micro is cheaper on listed token pricing. Granite 4.0 H Micro costs $0.02/1M input and $0.11/1M output tokens. LFM2.5-24B-A2B costs $0.05/1M input and $0.70/1M output tokens. Provider discounts or batch pricing can still change the final bill.
Is Granite 4.0 H Micro or LFM2.5-24B-A2B open source?
Granite 4.0 H Micro is listed under Apache 2.0. LFM2.5-24B-A2B is listed under LFM Open License v1.0. License labels affect whether you can self-host, redistribute weights, or rely only on hosted APIs, so confirm the upstream license before deployment.
Which is better for reasoning mode, Granite 4.0 H Micro or LFM2.5-24B-A2B?
LFM2.5-24B-A2B has the clearer documented reasoning mode signal in this comparison. If reasoning mode is mission-critical, validate it against the provider endpoint because model-level support and API-level exposure can differ.
Which is better for JSON / Tool use, Granite 4.0 H Micro or LFM2.5-24B-A2B?
Both Granite 4.0 H Micro and LFM2.5-24B-A2B expose JSON / Tool use. The better choice depends on benchmark fit, context budget, pricing, and whether your provider route exposes the same capability surface.
Where can I run Granite 4.0 H Micro and LFM2.5-24B-A2B?
Granite 4.0 H Micro is available on Cloudflare Workers AI and OpenRouter. LFM2.5-24B-A2B is available on Liquid AI, OpenRouter, and Vercel AI Gateway. Provider coverage can affect latency, region availability, compliance posture, and fallback options.
Continue comparing
Last reviewed: 2026-11-12. Data sourced from public model cards and provider documentation.