Granite 4.2 8B
A small dense 8B-parameter reasoning model from IBM for mathematics, code, multilingual dialogue and agent tasks with several steps. It can reason fully, with low effort or not at all, and supports 12 languages including English, German, Japanese, Arabic and Chinese. Text only, with a 128K-token context and tool calling.
Overview
131.1Ktokens
118Ktokens
0.07 USDper 1M tokens
0.3 USDper 1M tokens
31 August 2026
Start with Granite 4.2 8B
The model name is already filled in. Mint a key in the console, put it in an environment variable, and the call below runs as it stands — provided the organization’s wallet holds funds: a call with nothing to pay with answers 402.
curl https://api.kumorouter.com/v1/chat/completions \
-H "Authorization: Bearer $KUMO_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "granite-4-2-8b",
"max_tokens": 1024,
"messages": [
{
"role": "user",
"content": "Explain tokens in one line."
}
]
}'OpenAI — API parameters
Authorization: Bearer $KUMO_API_KEY
| Parameter | Type | Default / range | Description |
|---|---|---|---|
| model | string | Required | Published model identifier. |
| messages | array | 1–512 | Conversation messages in chronological order. |
| max_tokens | integer | ≥ 1 | Optional. A request stating neither max_tokens nor max_completion_tokens is answered under the surface default of 32768 output tokens. Model limits also apply. |
| temperature | number | 0–2 | Sampling randomness. Omission leaves the choice to the model. |
| top_p | number | 0–1 | Nucleus sampling probability mass. Omission leaves the choice to the model. |
| stop | array | ≤ 4 | Array of strings that stop generation. |
| max_completion_tokens | integer | ≥ 1 | Optional alternative to max_tokens. Stating both with DIFFERENT values is refused: a precedence rule would silently discard one of two numbers the caller deliberately wrote. |
| stream | boolean | false | Stream response events using Server-Sent Events. |
| stream_options | object | {include_usage: boolean} | Only with stream: true. include_usage adds a final usage event before [DONE]; its default is false. |
| tools | array | ≤ 128 | Function declarations. parameters carries the original JSON Schema object, up to 65536 bytes per tool, including nested objects and arrays. |
| tool_choice | string | object | auto | none | required | {type: "function", function: {name}} | auto leaves the choice to the model; none forbids a call while the declarations stay visible; required requires a call to some declared tool; the object form requires the named one. Omission leaves the choice to the model. |
One completion per request. The gateway rejects frequency_penalty, presence_penalty, seed, logprobs, top_logprobs and logit_bias: each of them would change the generation or the shape of the reply, and accepting one silently would misdescribe the answer you got. n is accepted only at 1 — the protocol default and exactly what the gateway does — and an n above one is rejected. user is accepted but is not forwarded to the model, not stored and not echoed back. The same applies to strict on a function declaration and name on a tool-role message: both are accepted and not forwarded. reasoning_effort is accepted and not forwarded either: no supplier request on this platform carries a reasoning budget, so the answer comes back at the default of the model itself whatever the field says. cache_control on a message, on a content part or on a replayed tool call is accepted and not forwarded either: the supplier wire this protocol is served by has no member for a prompt-cache anchor, so the cached-token counts come back as though the anchor had not been sent. messages[].content is accepted both as a string and as a list of parts: text parts are joined in order into one text, so a request written in parts and the same request written as a string are one request. A part of kind image_url, input_audio, file or refusal is rejected naming its own type: no multimodal capability is verified on this platform for any pair, and a replayed refusal would reach the supplier as ordinary assistant speech. The developer role is accepted and forwarded as system — it is the newer name this protocol gives the instruction role. stop is array-only; tool_choice is the string mode auto, none or required, or an object naming a declared tool. json_object is refused by name: promising valid JSON says nothing about its shape.
Identifiers
Capabilities
| Protocol and modality | Streaming | Tools | Structured output | Strict semantics | Structured streaming |
|---|---|---|---|---|---|
| Chat Completions text generation | Yes | Yes | Unverified | Unverified | Unverified |
A row describes one protocol-and-modality pair rather than the model as a whole: a capability is proved on a particular surface, and the answer beside it may differ.
How it compares
| Model | Context | Max output | Input $/M | Output $/M | Modalities |
|---|---|---|---|---|---|
| Granite 4.2 8B | 131.1K | 118K | 0.07 USD | 0.3 USD | text generation |
| Claude Sonnet 5 | 1M | 128K | 1.2 USD | 6 USD | text generation |
| Gemini 3.6 Flash | 1M | 65.5K | 0.4 USD | 1.8 USD | text generation |
| Gemini 3.8 Flash | 1M | 65.5K | 0.4 USD | 2.1 USD | text generation |
| Kimi K3 | 1M | 1M | 1.7 USD | 8.3 USD | text generation |
| Claude Opus 4.8 | 1M | 128K | 2.6 USD | 12.8 USD | text generation |
| Claude Fable 5 | 1M | 128K | 5.1 USD | 25.5 USD | text generation |
| Claude Fable 5.1 | 1M | 128K | 6 USD | 30 USD | text generation |
| Claude Haiku 4.5 | 200K | 64K | 0.6 USD | 3 USD | text generation |
| Claude Opus 4.5 | 200K | 64K | 2.6 USD | 12.8 USD | text generation |
| Claude Opus 4.6 | 1M | 128K | 2.6 USD | 12.8 USD | text generation |
| Claude Opus 4.7 | 1M | 128K | 2.6 USD | 12.8 USD | text generation |
| Claude Opus 5.5 | 1M | 128K | 2.2 USD | 11 USD | text generation |
| Claude Opus 5 | 1M | 128K | 2.6 USD | 12.8 USD | text generation |
| Claude Sonnet 5.5 | 1M | 128K | 1.1 USD | 5.5 USD | text generation |
| Claude Sonnet 4.5 | 200K | 64K | 1.5 USD | 7.7 USD | text generation |
| Claude Sonnet 4.6 | 1M | 128K | 1.5 USD | 7.7 USD | text generation |
| Gemini 3 Flash Preview | 1M | 65.5K | 0.3 USD | 1.5 USD | text generation |
| Gemini 3.1 Flash Lite | 1M | 65.5K | 0.2 USD | 0.9 USD | text generation |
| Gemini 3.1 Pro Low | 1M | 65.5K | 1.1 USD | 6.6 USD | text generation |
| Gemini 3.6 Flash High | 1M | 65.5K | 0.4 USD | 1.9 USD | text generation |
| Gemini 3.7 Flash High | 1M | 65.5K | 0.4 USD | 1.9 USD | text generation |
| Gemini 3.8 Flash High | 1M | 65.5K | 0.5 USD | 2.3 USD | text generation |
| Codex Auto Review | 1.1M | 128K | 3 USD | 18 USD | text generation |
| GPT-5.3 Codex Spark | 128K | 32K | 0.9 USD | 7.1 USD | text generation |
| GPT-5.5 | 1.1M | 128K | 2.6 USD | 15.3 USD | text generation |
| GPT-5.6 Luna | 1.1M | 128K | 0.1 USD | 0.6 USD | text generation |
| GPT-5.6 Sol | 1.1M | 128K | 2 USD | 10.2 USD | text generation |
| GPT-6.1 Sol | 1.1M | 128K | 1.1 USD | 5.5 USD | text generation |
| GPT-6 Astra | 1.1M | 128K | 6 USD | 30 USD | text generation |
| GPT-5.6 Terra | 1.1M | 128K | 1 USD | 6.1 USD | text generation |
| GPT-6 Sol | 1.1M | 128K | 1.1 USD | 5.5 USD | text generation |
| GPT-6 Luna | 1.1M | 128K | 0.06 USD | 0.3 USD | text generation |
| Grok 4.5 | 500K | 500K | 1 USD | 3.1 USD | text generation |
| Grok 4.6 | 500K | 500K | 1 USD | 3.1 USD | text generation |
| Grok 4.7 | 500K | 500K | 1.1 USD | 3.3 USD | text generation |
| GLM-5.2 | 1M | 131.1K | 0.7 USD | 2.2 USD | text generation |
| GLM-5.3 | 1M | 131.1K | 0.8 USD | 2.6 USD | text generation |
| Claude Haiku 5.5 | 1M | 128K | 0.06 USD | 0.3 USD | text generation |
| Mistral Large 4 | 1M | 262.1K | 0.8 USD | 2.4 USD | text generation |
| Mistral Medium 3.5 | 262.1K | 209.7K | 1.7 USD | 8.7 USD | text generation |
| Devstral 2 | 262.1K | 209.7K | 0.5 USD | 2.3 USD | text generation |
| DeepSeek V4.1 Flash | 1M | 393.2K | 0.3 USD | 1.4 USD | text generation |
| DeepSeek V4 Pro 0813 | 1M | 393.2K | 1.5 USD | 4.6 USD | text generation |
| DeepSeek V4 Flash | 1M | 943.7K | 0.1 USD | 0.2 USD | text generation |
| Qwen3.8 Max 0902 | 1M | 131.1K | 2.3 USD | 7 USD | text generation |
| Qwen3.8 Flash | 1M | 131.1K | 0.2 USD | 0.5 USD | text generation |
| Qwen3.8 27B | 1M | 235.9K | 0.05 USD | 2.7 USD | text generation |
| Qwen3.7 Plus | 1M | 131.1K | 0.4 USD | 1.5 USD | text generation |
| Qwen3 Coder Next | 262.1K | 235.9K | 0.1 USD | 0.9 USD | text generation |
| MiniMax M3 | 1M | 512K | 0.3 USD | 1.4 USD | text generation |
| MiniMax M2.7 | 204.8K | 176.9K | 0.3 USD | 1.4 USD | text generation |
| Kimi K2.7 Code | 262.1K | 235.9K | 0.8 USD | 3.9 USD | text generation |
| Kimi K2.6 | 262.1K | 235.9K | 0.8 USD | 4 USD | text generation |
| GLM-5.3 FlashX | 1M | 131.1K | 0.4 USD | 1.5 USD | text generation |
| MiMo V2.6 Pro | 1.1M | 131.1K | 0.5 USD | 1 USD | text generation |
| MiMo V2.6 Flash | 1.1M | 131.1K | 0.2 USD | 0.3 USD | text generation |
| Step 5 Preview | 1M | 64K | 1.2 USD | 3.1 USD | text generation |
| Step 3.7 Flash | 262.1K | 230.4K | 0.2 USD | 1.3 USD | text generation |
| Muse Glimmer 30B | 131.1K | 16.4K | 0.3 USD | 1.4 USD | text generation |
| Nemotron 3 Ultra | 262.1K | 182.5K | 0.7 USD | 2.8 USD | text generation |
| Nemotron 3 Super | 262.1K | 235.9K | 0.09 USD | 0.5 USD | text generation |
| Seed 2.1 Turbo | 262.1K | 235.9K | 0.6 USD | 2.9 USD | text generation |
| Seed 2.0 Code | 262.1K | 131.1K | 0.6 USD | 3.5 USD | text generation |
| Hy4 Preview | 1M | 64K | 1 USD | 2.9 USD | text generation |
| Command A+ | 192K | 64K | 0.3 USD | 1.7 USD | text generation |
| Inkling | 524.3K | 471.9K | 1.2 USD | 4.7 USD | text generation |
| LongCat 2.0 | 1M | 262.1K | 0.3 USD | 1.4 USD | text generation |
| Nova 2 Lite | 1M | 65.5K | 0.3 USD | 2.9 USD | text generation |
| Trinity Large Thinking | 262.1K | 80K | 0.3 USD | 0.9 USD | text generation |
| Solar Pro 4 | 524.3K | 131.1K | 0.1 USD | 0.4 USD | text generation |
| Aion 3.5 | 262.1K | 32.8K | 3.5 USD | 7 USD | text generation |
| Laguna S 2.1 | 1M | 131.1K | 0.1 USD | 0.2 USD | text generation |
| Mercury 2.5 | 260K | 65.5K | 0.05 USD | 0.2 USD | text generation |
| Llama 4 Maverick | 1M | 16.4K | 0.2 USD | 0.8 USD | text generation |
Prices
| Dimension | Rate |
|---|---|
| Input tokens | 0.07 USD per 1M tokens |
| Cache write | 0.07 USD per 1M tokens |
| Cache read | 0.02 USD per 1M tokens |
| Output tokens | 0.3 USD per 1M tokens |
Integrations
Every link opens its own guide with this model already chosen.