Gemini 4 Carbon

Google reasoning model for software engineering, agents, multimodal knowledge work, mathematics and science. Shared 1048576-token context; API-reported synchronous output limit 32768.

Input / output priceFree: $0.00 / $0.00 per 1M
Context: 1M tokens
Released: Nov 10, 2026

Try Respan for free

Agent setup

Paste into Claude Code, Cursor, or Codex

Help me set up the Respan gateway: fetch and follow https://www.respan.ai/docs/agent-setup.md. Use model "google_gemini_ai/gemini-4-carbon:free".

from openai import OpenAI

client = OpenAI(

    base_url="https://api.respan.ai/api/",

    api_key="YOUR_RESPAN_API_KEY",

)

response = client.chat.completions.create(

    model="google_gemini_ai/gemini-4-carbon:free",

    messages=[{"role": "user", "content": "Hello!"}],

)

print(response.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({

  baseURL: "https://api.respan.ai/api/",

  apiKey: "YOUR_RESPAN_API_KEY",

});

const response = await client.chat.completions.create({

  model: "google_gemini_ai/gemini-4-carbon:free",

  messages: [{ role: "user", content: "Hello!" }],

});

console.log(response.choices[0].message.content);

curl https://api.respan.ai/api/chat/completions \

  -H "Authorization: Bearer YOUR_RESPAN_API_KEY" \

  -H "Content-Type: application/json" \

  -d '{

    "model": "google_gemini_ai/gemini-4-carbon:free",

    "messages": [{"role": "user", "content": "Hello!"}]

}'

Read docs

Providers

6 routes from 4 providers, cheapest first, OpenRouter last, priced per 1M tokens after gateway discounts.

Provider	Context	Input	Output	Cache read	Capabilities
Google Gemini AIFree`google_gemini_ai/gemini-4-carbon:free`	1M	$0.00/M	$0.00/M	$0.00/M
RespanFree`respan/gemini-4-carbon:free`	1M	$0.00/M	$0.00/M	$0.00/M
Google Gemini AI50% off`google_gemini_ai/gemini-4-carbon`	1M	~~$4.00/M~~$2.00/M	~~$20.00/M~~$10.00/M	~~$0.20/M~~$0.10/M
Perplexity`perplexity/google/gemini-4-carbon`	—	$2.00/M	$10.00/M	$0.10/M	—
Respan50% off`respan/gemini-4-carbon`	1M	~~$4.00/M~~$2.00/M	~~$20.00/M~~$10.00/M	~~$0.20/M~~$0.10/M
OpenRouter`openrouter/google/gemini-4-carbon`	1M	$2.00/M	$10.00/M	$0.10/M

Is a provider down? Perplexity status, OpenRouter status

Pricing

Averaged across 6 routes from 4 providers, in USD per 1M tokens after gateway discounts.

Average input: $1.33 per 1M
Average output: $6.67 per 1M
Average cache read: $0.07 per 1M
Average cache write: $0.09 per 1M

Cost by volume

Total cost from 0 to 1T tokens on the cheapest and most expensive provider.

Highest$2.00/MGoogle Gemini AI, Perplexity, Respan, and OpenRouterLowest$0.00/MGoogle Gemini AI and Respan

Throughput and latency

30-day averages from requests through the Respan gateway. Under each figure, every model on the gateway by its average, this one in blue.

Output speed

Higher is better

412tokens/sacross providers

Faster than 89% of the other 430 models on the gateway

Time to first token

Lower is better

58sacross providers

Faster than 0% of the other 119 models on the gateway

Response latency

Lower is better

63sacross providers

Faster than 3% of the other 492 models on the gateway

Benchmarks

The Respan Index plus 25 published scores from 8 sources, each linked to where it was published.

76.3/ 100

Respan Index#2 of 169

169 ranked models. The shaded band is this model's 95% interval, 71.7 to 80.2. Claude Opus 5.5 falls inside it, so the gap is within the margin of error.

Published scores

Benchmark	Score	Source	Published
CWE-benchGoogle	71%	Google	Nov 10, 2026
Terminal-Bench 4.0 (AA)Google	65.2%	Google	Nov 10, 2026
Terminal-Bench 4.0Vals AI	63.1%	Vals AI	Nov 11, 2026
FrontierSWE 2Google	61.2%	Google	Nov 10, 2026
FrontierSWE V2Proximal Labs	61.2%	Proximal Labs	Nov 11, 2026

Benchmark	Score	Source	Published
APEX-AgentsMercor	84.2%	Mercor	Nov 11, 2026
OSWorld-2.1 (Offline subset)Google	81.5%	Google	Nov 10, 2026
Agents' Last Exam v1.1Google	77.4%	Google	Nov 10, 2026
Vending-Bench 2Andon Labs	16484.21 · #1 of 65	Andon Labs	Nov 11, 2026

Benchmark	Score	Source	Published
GPQA DiamondEpoch AI	96.5%	Epoch AI	Nov 11, 2026
GraphWalks 256k-1M BFS (F1)Google	96.2%	Google	Nov 10, 2026
SimpleBenchSimpleBench	92.6%	SimpleBench	Nov 10, 2026
EBR-benchEpoch AI	87.6%	Epoch AI	Nov 11, 2026
Chess PuzzlesEpoch AI	78%	Epoch AI	Nov 11, 2026
Mystery Game PuzzlesEpoch AI	73%	Epoch AI	Nov 11, 2026

Benchmark	Score	Source	Published
OTIS Mock AIME 2024-2025Epoch AI	100%	Epoch AI	Nov 11, 2026
FrontierMath Tier 4 v2Epoch AI	95.1%	Epoch AI	Nov 11, 2026
FrontierMath Tiers 1-3 v2Epoch AI	90.5%	Epoch AI	Nov 11, 2026

Benchmark	Score	Source	Published
SimpleQA VerifiedEpoch AI	77.7%	Epoch AI	Nov 11, 2026

Benchmark	Score	Source	Published
Furniture AssemblyEpoch AI	77.5%	Epoch AI	Nov 11, 2026

Benchmark	Score	Source	Published
LMArena WebDevLMArena	1768 Elo · #4 of 118	LMArena	Nov 10, 2026
LMArena EloLMArena	1526 Elo · #1 of 176	LMArena	Nov 9, 2026
LMArena AgentLMArena	0.1663 · #2 of 60	LMArena	Nov 8, 2026

Benchmark	Score	Source	Published
SEC-bench ProGoogle	88.2%	Google	Nov 10, 2026
AA Index v5.0Google	59	Google	Nov 10, 2026

How the Respan Index works

How Gemini compares: Claude vs Gemini, Gemini vs GPT, Gemini vs Grok

Apps

Use Gemini 4 Carbon in the tools you already run. Each guide points the app at the Respan gateway.

Setup guides for more apps are in the integration docs.

Gemini 4 Carbon FAQ

Short answers, built from the data on this page.

Gemini 4 Carbon is a model from Google Gemini AI, released on Nov 10, 2026. Google reasoning model for software engineering, agents, multimodal knowledge work, mathematics and science. Shared 1048576-token context; API-reported synchronous output limit 32768.

On the Respan gateway, Gemini 4 Carbon costs from $0.00 per 1M input tokens and $0.00 per 1M output tokens on Google Gemini AI and Respan, after gateway discounts. Averaged across its 6 routes from 4 providers, it costs $1.33 per 1M input tokens and $6.67 per 1M output tokens.

Gemini 4 Carbon is available through 6 routes from 4 providers: Google Gemini AI, Respan, Perplexity, and OpenRouter.

Gemini 4 Carbon accepts up to 1M tokens of context on its largest provider route.

Point any OpenAI-compatible SDK at https://api.respan.ai/api/ with your Respan API key and set the model to "google_gemini_ai/gemini-4-carbon:free". To set it up with a coding agent, copy the agent setup prompt at the top of this page.

Gemini 4 Carbon scores 76.3 out of 100 on the Respan Index v1.1, ranking #2 of 169 models. By category: Coding 69.4, Agents 83.4, Reasoning 84.7, Knowledge & language 77.8, and Human preference 61.0. The Respan Index combines independent and lab-reported benchmark results into one score, using benchmark data collected by Respan.

Gemini 4 Carbon has 25 published benchmark scores, including CWE-bench 71%, APEX-Agents 84.2%, and GPQA Diamond 96.5%. Each score links to where it was published; the data is collected by Respan.