MODEL PROFILE

gemini-4-argon-thinking-low

Google · gemini-4-argon-thinking-low

The special offer group model is implemented by partially eligible accounts, using specific parameter requests to invoke Gemini 3.8 Flash. It has been confirmed that upstream traffic is indeed routed to Gemini 4. These are different checkpoints within Gemini 4; reportedly, b performs slightly better than a, but b currently has a lower level of reasoning. The exact IQ should be assessed through actual customer testing.

Text generation对话识图Reasoning
Context windowNot disclosed
Input → OutputText、Image → Text
Model weightsNot verified
Profile updated2026/11/2 19:17:09
OVERVIEW

Model overview

The special offer group model is implemented by partially eligible accounts, using specific parameter requests to invoke Gemini 3.8 Flash. It has been confirmed that upstream traffic is indeed routed to Gemini 4. These are different checkpoints within Gemini 4; reportedly, b performs slightly better than a, but b currently has a lower level of reasoning. The exact IQ should be assessed through actual customer testing.

The summary below is compiled from sources by an AI model, not our own testing. Verify selection decisions against original sources and your workload.

Key features

  • Supports multimodal input of text and images
  • The actual backend routes to the Gemini 4 model (not Gemini 3.8 Flash)
  • A special access channel open to certain eligible accounts

Limitations & considerations

  • The reasoning level is explicitly labeled as ‘low,’ and the reasoning capability is limited.
  • Performance depends on users’ actual testing and lacks objective benchmark metrics.
  • Access permissions are subject to dual restrictions: account eligibility and specific parameters.
PRICING & COST

Source prices & cost analysis

Prices vary by channel, region and configuration. Channel quotes are listed independently and are not treated as a single official provider price.

Source / channelInputOutputBilling unitCollected at
Apilio · gemini-t3 ↗Not verifiedNot verified渠道倍率待核验2026/11/2 06:25:12
Apilio · gemini优质 ↗Not verifiedNot verified渠道倍率待核验2026/11/2 06:25:12
Apilio · origin ↗Not verifiedNot verified渠道倍率待核验2026/11/2 06:25:12

Estimate monthly base token costs

—

Estimates exclude cache discounts, tiered pricing, multimodal surcharges, web search, taxes and additional reasoning output.

INDEPENDENT BENCHMARKS

Public benchmarks & run configuration

暂无已采集的可核验公开评分。模型未上榜不代表能力较低。

View full indexed rankings →

Frequently asked questions

Is this model really Gemini 4?

According to the data, the upstream detection confirms that the request is actually routed to Gemini 4, even though the call must be accompanied by specific parameters and nominally points to Gemini 3.8 Flash.

Why is it called 'thinking-low'?

The 'thinking-low' in the name indicates that this checkpoint currently has a lower enabled thinking level, meaning its reasoning depth or chain-of-thought capability is deliberately limited.

Who can use this model?

It is available only to accounts with specific qualifications, is not publicly accessible, and must meet the required parameter configuration.

Do open weights permit free commercial use?

Not necessarily. Check the version license, commercial conditions and usage policies. Self-hosting incurs compute and maintenance costs.

Are these real-time billing prices?

These are timestamped source snapshots, not live billing prices. Check the channel for current versions, regions, discounts and billing methods.

Sources & traceability

许可证:来源未披露,须进一步核验。模型介绍与第三方评测均以原来源的具体版本和方法为准。