Gemini 4 Argon is the Gemini 4 model Google announced on September 30, 2026, with a 1,048,576-token context window shared by input, thinking, and output, and outputs of up to 1M tokens through continuation. Call it through EvoLink with the API key you already use for other models: $1.800 per 1M input tokens and $9.000 per 1M output tokens while Google's introductory pricing is in effect.
Google·Text Generation·Available
From $1.800 / 1M input tokens$4.000 official priceIntro pricing-55%
Long outputPrompt cachingGoogle SearchChat + Gemini API
Production routeThis rate reflects platform-side availability — only confirmed server errors (HTTP 500 / empty response) count as failures. User-side issues (content moderation, invalid params, cancellation) plus rate limits, timeouts and auth errors are excluded. Before real traffic arrives, empty buckets may display as available.Live
Google's Gemini 4 model, announced on September 30, 2026, with an output limit of 1M tokens. Input, cached input, and output are metered separately; current rates follow Google's introductory pricing.
Gemini 4 Argon
Google's Gemini 4 model
Selected
Model ID
gemini-4-argon
Best for
Requests that need a long answer, and workloads that resend the same long prompt and can use the cached-input rate. Compare results and total cost on your own tasks before moving production traffic.
Input
$1.800 / 1M-55%
122.4 cr / 1M$4.000official price
Cache read
$0.092 / 1M-54%
6.2 cr / 1M$0.200official price
Output
$9.000 / 1M-55%
612 cr / 1M$20.000official price
All rates are per 1M tokens, shown in USD and credits, and reflect your account's current pricing.
Gemini 4 Argon pricing
Estimate what one Gemini 4 Argon request costs before you integrate. The calculator uses your account's current rates, with official pricing as a reference.
Introductory pricing — the rates below follow Google's introductory price for Gemini 4 Argon ($2.000 input / $10.000 output per million tokens) and run through November 30, 2026 (UTC+8). From December 1, 2026, EvoLink's usual discount applies to Google's standard price ($4.000 input / $20.000 output per million tokens).
Request calculator
Enter the token mix for one request and the number of Google Search queries.
Estimated request cost
Gemini 4 Argon
USD$0.0046
Credits0.3073
Official estimate $0.011 · save $0.0056 (55%)
Input tokens0.1224 cr
Cache read tokens0.0013 cr
Output tokens0.1836 cr
Google Search0 cr
Minimum charge: 0.01 credits per request.
Budget guide
Approximate requests using the current token mix and search queries.
Google Search is billed per search query, and one request may run several queries. Search fees are added separately after the token minimum charge.
Google Search$0.014/ query·0.952 cr / query
Gemini 4 Argon API
Gemini 4 Argon is the Gemini 4 model Google announced on September 30, 2026, with a 1,048,576-token context window shared by input, thinking, and output, and outputs of up to 1M tokens through continuation. Call it through EvoLink with the API key you already use for other models: $1.800 per 1M input tokens and $9.000 per 1M output tokens while Google's introductory pricing is in effect.
Gemini 4 Argon is served on EvoLink under the model ID gemini-4-argon through Chat Completions · Gemini native API, with the same API key and balance you use for every other model. It offers a 1.05M context window and up to 1.05M output tokens, plus prompt caching, server-side tools.
Gemini 4 Argon specs and capabilities
Numbers come from the EvoLink route configuration; capabilities are what the API exposes today.
Context window
1.05M tokens
Max output
1.05M tokens
Input
Text + images
Output
Text · JSON (structured output) · tool calls
Reasoning
Configurable reasoning effort
Tool use
Function calling with multi-step tool sequences
Prompt caching
Automatic cache reads at a lower rate
Server-side tools
Google Search, billed per search query
Protocols
Chat Completions · Gemini native API
Model ID
gemini-4-argon
What Gemini 4 Argon is a good fit for
These starting points come from what Google has published so far: the output limit and the pricing. Check them against your own workload before you move traffic.
Long outputs with Gemini 4 Argon
A single request returns up to 65,536 thinking and output tokens (32,768 without streaming); longer results, up to 1M tokens, continue across requests on the native API. That limit is capacity, not a guarantee of a complete or correct answer. For a numbered technical question bank, request IDs and a fixed answer schema, verify every expected ID, sample derivations and run generated code tests. If output truncates or answers contradict each other, split into smaller batches and recheck the combined result.
Repeated long prompts with cached input
Google prices cached input at 95% off the input rate. When requests reuse the same long prefix — a system prompt, reference documents, or earlier turns — the cached part is billed at $0.092 per 1M tokens on EvoLink instead of $1.800.
Side-by-side evaluation on one API key
Gemini 4 Argon uses the same EvoLink key, balance, and endpoints as the models you already call. Replay the same tasks against the model you run today and compare accepted results and total cost before you switch a workload.
Two ways to use Gemini 4 Argon: EvoLink API or Agent
Use the EvoLink API for product backends and batch jobs, or call Gemini 4 Argon from Codex, Claude, or Gemini for coding and analysis workflows. Both paths share the same EvoLink API key, balance, model ID, and request history.
Option 1
Integrate with the EvoLink API
Best for: product backends, batch jobs, automated pipelines
Send Gemini native generateContent requests to EvoLink and control the model ID, system instruction, output budget, tools, and structured output.
1Create an EvoLink API key in the console
2Call the generateContent endpoint on the EvoLink base URL with the model ID shown above in the request path
3Send one representative request and read usageMetadata for input, cached, and output tokens
4Set maxOutputTokens and retries per task; keep function-call parts and their results across turns
Best for: coding, review, and analysis tasks in Codex, Claude, and Gemini
Give the Agent the task, the inputs to include, and the acceptance criteria. It assembles the request, calls Gemini 4 Argon through EvoLink, and returns the answer with token usage.
1Set EVOLINK_API_KEY in your local environment; never put it in code or a prompt
2Describe the task, the inputs to include, and the expected output format
3Ask the Agent to call Gemini 4 Argon through EvoLink and show the request before sending
4Let the Agent report the answer, token usage, and any error body
Gemini 4 Argon API code example and error handling
This example shows the shortest runnable request: a Gemini native generateContent call with a system instruction, a user message, and an output budget. Open the API tab for the complete parameter and response reference.
# Gemini native API
curl -X POST "https://direct.evolink.ai/v1beta/models/gemini-4-argon:generateContent" \
-H "Authorization: Bearer $EVOLINK_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"systemInstruction": { "parts": [{ "text": "You are a precise technical writer." }] },
"contents": [
{ "role": "user", "parts": [{ "text": "Write the migration guide for the change described below:\n<change notes>" }] }
],
"generationConfig": {
"maxOutputTokens": 8192,
"thinkingConfig": { "thinkingLevel": "high" }
}
}'
# thinkingLevel accepts low, medium, or high. For streaming, call
# streamGenerateContent?alt=sse with the same body.
# Output tokens, including thinking tokens, are billed at the output rate,
# so size maxOutputTokens for the task. The response includes
# candidates[0].content.parts and usageMetadata (prompt, cached, thinking,
# and output token counts).
Invalid request or unsupported parameter
Check the model ID, the contents array, and parameter ranges against the API reference; remove fields this route does not support.
Authentication or balance issue
Check the Authorization bearer token and confirm the available balance in the console.
Context length exceeded
Prompt tokens exceed the Gemini 4 Argon context window. Trim or retrieve only the relevant evidence and reuse cached prefixes.
Rate limited (429)
Back off and retry with jitter; batch or queue requests instead of sending parallel bursts.
Content or tool call rejected
Review sensitive content, malformed tool-call arguments, and JSON schema mismatches before retrying.
Gemini 4 Argon API integration checks
Keep one EvoLink client and API key. These four steps cover what is specific to Gemini 4 Argon.
01
Set the model ID
For Chat Completions, set model to "gemini-4-argon" in the request body. For the native API, use that ID in /v1beta/models/{model}:generateContent. "gemini-4" is this page URL, not a model ID; the two protocols have different request shapes.
Required
02
Budget the output
A single request returns at most 65,536 thinking and output tokens, or 32,768 without streaming. In Chat Completions, max_tokens limits one response, and a response cut at the limit ends with finish_reason "length". On the native API, set maxOutputTokens above the per-request limit to allow continuation. Separately cap the whole Agent task: for example, 8,000 output tokens per call, at most 3 calls and an explicit application spending ceiling. These are illustrative application limits, not model defaults. Stop before a retry would exceed the budget, and include input, output and tool fees.
Cost control
03
Reuse stable prefixes
Keep reusable prefixes stable, then verify actual cache hits rather than assuming every repeated prompt is discounted. Chat Completions reports usage; the native API reports usageMetadata. Read the documented cached-token fields for the selected protocol and reconcile them with billing.
Cost control
04
Compare before you switch
Record task ID, model/route, request settings, acceptance result, retries, billed usage, elapsed time and review minutes for both models. Use the same inputs and acceptance rules. Keep a tested, still-available rollback route; see the GPT-6 Astra comparison for protocol differences and workload-specific checks.
Evaluate
Cached input at a lower rate
Google prices cached input at 95% off the input rate. EvoLink cached input is $0.092 per 1M tokens, against Google’s standard $0.200. Caching is automatic and applies only to shared prefixes of at least 6,144 tokens. Verify actual cache hits in the selected protocol’s usage fields and billing; repeated text alone does not prove a cache hit.
One 1,048,576-token window; up to 65,536 tokens per request
Input, thinking, and output share one 1,048,576-token context window. Google raised the maximum output from 64K to 1,048,576 tokens, but a single request returns at most 65,536 thinking and output tokens (32,768 without streaming). To go further on the native API, set maxOutputTokens above that limit: a response that ends with the finish reason CONTINUATION carries a continuationToken, and resending the same request with that token continues the output until it ends with STOP. Chat Completions cannot continue; a response cut at the limit ends with finish_reason "length".
Gemini 4 Argon vs Gemini 3.8 Flash and Gemini 3.1 Pro
EvoLink
Prices are the current EvoLink rates for each model. Use the table to shortlist, then compare results and total cost per task on your own workload.
Is the Gemini 4 Argon API available through EvoLink?
Yes. Gemini 4 Argon is available on EvoLink. Use your existing EvoLink API key and send model "gemini-4-argon".
What model ID should I use for the Gemini 4 Argon API?
For Chat Completions, set model to "gemini-4-argon" in the request body. For the native API, use that ID in /v1beta/models/{model}:generateContent. "gemini-4" is this page URL, not a model ID; the two protocols have different request shapes.
How much does the Gemini 4 Argon API cost on EvoLink?
$1.800 per 1M input tokens, $9.000 per 1M output tokens, and $0.092 per 1M cached input tokens. The discounts shown on this page are measured against Google's standard price of $4.000 input and $20.000 output per 1M tokens, not the introductory rate.
What is Google's introductory pricing for Gemini 4 Argon, and when does it end?
Google launched Gemini 4 Argon at $2.000 per 1M input tokens and $10.000 per 1M output tokens, with cached input at 95% off the input rate ($0.100). Google's introductory pricing runs through November 30, 2026; from December 1, 2026, its prices are $4.000 and $20.000, with cached input at $0.200. EvoLink's current rates follow the introductory price through November 30, 2026 (UTC+8); from December 1, 2026, our usual discount applies to Google's standard price.
Does Gemini 4 Argon's 1M-token limit refer to output or context?
Both. Gemini 4 Argon has one 1,048,576-token context window shared by input, thinking, and output, and Google raised the maximum output from 64K to the same 1,048,576 tokens. A single request returns at most 65,536 thinking and output tokens (32,768 without streaming). On the native API, longer outputs continue across requests: when a response ends with the finish reason CONTINUATION, resend the same request with its continuationToken until it ends with STOP. These limits apply to one turn, not to a multi-turn Agent task. Add a separate application cap for all calls, retries and tool fees.
How is cached input billed on Gemini 4 Argon?
Cached input is billed at $0.092 per 1M cached tokens on EvoLink. Caching is automatic and applies only to shared prefixes of at least 6,144 tokens. Keep reusable prefixes stable, then verify actual cache hits rather than assuming every repeated prompt is discounted. Chat Completions reports usage; the native API reports usageMetadata. Read the documented cached-token fields for the selected protocol and reconcile them with billing.
How is Google Search billed with Gemini 4 Argon?
Google Search runs through the Gemini native API (the googleSearch tool) and costs $0.014 per search query, in addition to token charges. One request may run several queries.
Which protocols work with Gemini 4 Argon?
The OpenAI-compatible Chat Completions endpoint and the Gemini native API, both on direct.evolink.ai with the same EvoLink API key and the same model ID.
How should I compare Gemini 4 Argon with the model I use now?
Reuse your EvoLink API key and select the documented protocol and model ID for each route. Check tool formats and response handling, then replay representative tasks and compare accepted results, retries, output tokens, and total cost before moving production traffic.
