--- title: "IBM: Granite 4.2 8B — price, capability and what beats it · Undominated.ai" canonical: https://undominated.ai/models/ibm-granite__granite-4.2-8b/ description: "IBM: Granite 4.2 8B: $0.060/M in, $0.250/M out. Independent capability scores, and the models that are both better and cheaper." --- # IBM: Granite 4.2 8B — price, capability and what beats it · Undominated.ai > IBM: Granite 4.2 8B: $0.060/M in, $0.250/M out. Independent capability scores, and the models that are both better and cheaper. MODEL PROOF # Granite 4.2 8B IBM · 31 Aug 2026 Balanced Summarise Chat Code gen Agentic TRADE-OFF ## A stored alternative scores higher and costs less, with capability losses. **$0.107** / 1M tokens Balanced · 3 tokens in per 1 out Qwen3 30B A3B Instruct 2507 · 66.8 points higher independent score · 21% lower measured price - 118K → 32K max output - no extended reasoning Independent LMArena score 1,316.8 ± 11.38 · 3,087 votes Context window 131.07K Maximum output 117.96K Input modalities text Output modalities text Published input price $0.060 / 1M tokens Published output price $0.250 / 1M tokens Pricing kind fixed Evidence as of 22 Sept 2026 openrouter/models LMArena · cc-by-4.0 Copy link Share Download landscape card MODEL MONUMENT ## Save or share this proof IBM: Granite 4.2 8B. Balanced workload. Alternative has a trade-off. Evidence as of 22 Sept 2026. Download landscape card — 1200 by 630 pixels Download square card — 1080 by 1080 pixels Copy landscape image Share landscape card $0.060/M in, $0.250/M out in this catalogue, as of the fetch date on this page. That is the API row, not a subscription. Qwen3 30B A3B Instruct 2507 scores higher and costs less — but you would give something up. ### Every price dimension | Input | $0.060 /M | | --- | --- | | Output | $0.250 /M | | Cached input | $0.015 /M | ### Who sells it The same weights, different shops. Cheapest is not like-for-like when serving precision differs. | Seller | Input | Output | Precision | Uptime (1d) | | --- | --- | --- | --- | --- | | Seller DeepInfra | Input $0.060 | Output $0.250 | Precision bf16 | Uptime (1d) 99.99% | | Seller CoreWeave | Input $0.100 | Output $0.150 | Precision bf16 | Uptime (1d) 99.72% | ### Independent scores | LMArena Elo | 1316.8 default | | --- | --- | LMArena Elo under CC BY 4.0. Absence of another board is not a score of zero. ### Capability | Context window | 131K | | --- | --- | | Max output | 118K | | Input modes | text | | Tool use | yes | | Reasoning | optional | ### Provenance | Price source | openrouter.ai | | --- | --- | | Fetched | 2026-09-22 | | Quality data | verified | Granite 4.2 8B is a dense reasoning model from IBM. It is suited for mathematics, code generation, multilingual dialogue, and agentic workflows that need multi-step reasoning. It supports full, low-effort,... ## Questions this page answers ### What does Granite 4.2 8B cost? $0.060/M in, $0.250/M out in this catalogue, as of the fetch date on this page. That is the API row, not a subscription. ### Does Granite 4.2 8B have an independent quality score? Granite 4.2 8B has an LMArena score in this catalogue. ### What beats Granite 4.2 8B? Qwen3 30B A3B Instruct 2507 scores higher and costs less — but you would give something up. ### Does this page use Artificial Analysis scores for Granite 4.2 8B? No. Artificial Analysis figures are not published here. Quality on this page is LMArena Elo where a score exists; otherwise the row is unrated. [Check whether this is still the one to buy](/check/ibm-granite__granite-4.2-8b/) [Cheaper alternatives](/alternatives/ibm-granite__granite-4.2-8b/) ## Related - Higher score and cheaper, with a named trade [Qwen3 30B A3B Instruct 2507](/models/qwen__qwen3-30b-a3b-instruct-2507/) - Nearby on the capability axis [Qwen3 30B A3B](/models/qwen__qwen3-30b-a3b/) - Same vendor [Granite 4.0 Micro](/models/ibm-granite__granite-4.0-h-micro/) - IBM models [IBM](/providers/ibm/) - Where this model appears [spreads](/spreads/) - Nearby on the capability axis [GPT-4o (2024-05-13)](/models/openai__gpt-4o-2024-05-13/) - Where this model appears [significance](/significance/) - Nearby on the capability axis [o3 Mini](/models/openai__o3-mini/) - Where this model appears [cheapest-at](/cheapest-at/) ## Badge The verdict as one SVG, for a README or a docs page. It states this model's dominance status at the balanced workload on the LMArena lens, with the date it was computed, and it is rebuilt with the catalogue — so it changes when the verdict changes, including to one you would rather it did not. Markdown Copy [![Granite 4.2 8B — Undominated.ai dominance verdict](https://undominated.ai/badge/ibm-granite__granite-4.2-8b.svg)](https://undominated.ai/models/ibm-granite__granite-4.2-8b/) HTML Copy Granite 4.2 8B — Undominated.ai dominance verdict Direct file: [https://undominated.ai/badge/ibm-granite__granite-4.2-8b.svg](https://undominated.ai/badge/ibm-granite__granite-4.2-8b.svg) — a static SVG, written by scripts/build-badges.mjs on every build, so the copy you embed is never older than the last deploy.