borg Download

Model maker

Google models in Borg

Google develops Gemini models across Pro, Flash, and Flash-Lite tiers. Borg includes several Gemini generations and gateway effort variants.

Models carried
12
Confirmed reasoning models
12
Maximum published context
1,048,576 tokens Vendor source

Explore Google models

Search by model name, lane id or use case.

ModelContextReasoningReleaseBest forBorg speedBench
Gemini 2.5 Flash1,048,576 tokens Vendor sourceYes2025-04-17 Vendor sourcemultimodal document and video summaries142.4 output tokens/s (whole turn)—
Gemini 2.5 Flash-Lite1,048,576 tokens Vendor sourceYes2025-06-17 Vendor sourcecheap high-volume classification62.2 output tokens/s (whole turn)—
Gemini 2.5 Pro1,048,576 tokens Vendor sourceYes2025-03-25 Vendor sourcelong-repository code analysis95.3 output tokens/s (whole turn)—
Gemini 3 Flash1,048,576 tokens Vendor sourceYes2025-12-17 Vendor sourcevisual and spatial reasoning107.6 output tokens/s (whole turn)—
Gemini 3.1 Flash Image131,072 tokens Vendor sourceYes2026-02-26 Vendor sourcesearch-grounded image creation——
Gemini 3.1 Flash-Lite1,048,576 tokens Vendor sourceYes2026-03-03 Vendor sourcelarge-scale translation112.9 output tokens/s (whole turn)—
Gemini 3.1 Pro1,048,576 tokens Vendor sourceYes2026-02-19 Vendor sourcecomplex software engineering——
Gemini 3.5 Flash1,048,576 tokens Vendor sourceYes2026-05-19 Vendor sourceiterative coding agents——
Gemini 3.5 Flash-Lite1,048,576 tokens Vendor sourceYes2026-07-21 Vendor sourcehigh-throughput document parsing59.2 output tokens/s (whole turn)—
Gemini 3.6 Flash1,048,576 tokens Vendor sourceYes2026-07-21 Vendor sourcerapid agentic coding loops56.8 output tokens/s (whole turn)—
Gemini 3.7 Flash1,048,576 tokens Vendor sourceYes2026-08-13 Vendor sourcevideo-informed agent workflows72.8 output tokens/s (whole turn)—
Gemini 3.8 Flash1,048,576 tokens Vendor sourceYes2026-09-02 Vendor sourcelong-horizon coding agents29.2 output tokens/s (whole turn)—

Models in detail

These are the Google models currently listed as live in the Borg directory. The descriptions below summarize each model's published use cases; the linked model pages give the full account. A lane id is the value you use to select a model directly in the Borg gateway, not a separate model name. Check the linked pages for context windows, reasoning support and any vendor API list prices before choosing a lane for your work.

Gemini 2.5 Flash

Gemini 2.5 Flash is the middle member of Google's 2.5 family: less costly than Pro and more capable for involved work than Flash-Lite.

Best for Gemini 2.5 Flash:

  • multimodal document and video summaries
  • agentic coding with adjustable thinking
  • high-volume reasoning requests

Borg lane ids: antigravity/gemini-2.5-flash, antigravity/gemini-2.5-flash-thinking, google/gemini-2.5-flash. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 2.5 Flash-Lite

Gemini 2.5 Flash-Lite is the smallest, fastest and lowest-cost option in the 2.5 lineup, below Flash and substantially lighter than Pro.

Best for Gemini 2.5 Flash-Lite:

  • cheap high-volume classification
  • structured extraction from PDFs
  • low-latency message triage

Borg lane ids: antigravity/gemini-2.5-flash-lite, google/gemini-2.5-flash-lite. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 2.5 Pro

Gemini 2.5 Pro is the largest reasoning-oriented member of Google's 2.5 generation, aimed at problems where Flash and Flash-Lite may sacrifice depth for speed or price.

Best for Gemini 2.5 Pro:

  • long-repository code analysis
  • multi-step STEM reasoning
  • cross-document synthesis

Borg lane ids: antigravity/gemini-2.5-pro, google/gemini-2.5-pro, google/gemini-2.5-pro-preview. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3 Flash

Gemini 3 Flash is Google's initial Flash model in the third generation, built on the Gemini 3 Pro reasoning foundation but positioned for quicker and cheaper use.

Best for Gemini 3 Flash:

  • visual and spatial reasoning
  • interactive coding agents
  • multimodal document analysis

Borg lane ids: antigravity/gemini-3-flash, google/gemini-3-flash-preview. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3.1 Flash Image

Gemini 3.1 Flash Image, also called Nano Banana 2, is the image-producing branch of Google's Gemini 3 family, not a general text-only replacement for Flash or Pro.

Best for Gemini 3.1 Flash Image:

  • search-grounded image creation
  • conversational image editing
  • infographics with rendered text
  • multi-resolution visual assets

Borg lane ids: antigravity/gemini-3.1-flash-image, google/gemini-3.1-flash-image, google/gemini-3.1-flash-image-preview. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3.1 Flash-Lite

Gemini 3.1 Flash-Lite brings the Lite tier into Google's third-generation Gemini family.

Best for Gemini 3.1 Flash-Lite:

  • large-scale translation
  • support-ticket classification
  • lightweight subagent tool calls

Borg lane ids: antigravity/gemini-3.1-flash-lite, google/gemini-3.1-flash-lite, google/gemini-3.1-flash-lite-preview. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3.1 Pro

Gemini 3.1 Pro is the heavier reasoning tier of Google's Gemini 3 lineup, refining the earlier 3 Pro for better thinking, token efficiency and reliable multi-step tool use.

Best for Gemini 3.1 Pro:

  • complex software engineering
  • multi-step tool-driven agents
  • long-context multimodal research

Borg lane ids: antigravity/gemini-3.1-pro-high, antigravity/gemini-3.1-pro-low, google/gemini-3.1-pro-preview. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3.5 Flash

Gemini 3.5 Flash moves the Flash line toward sustained agentic work instead of simply answering isolated prompts.

Best for Gemini 3.5 Flash:

  • iterative coding agents
  • long-horizon subagent workflows
  • multimodal investigation

Borg lane ids: google/gemini-3.5-flash. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3.5 Flash-Lite

Gemini 3.5 Flash-Lite is the economical counterpart to 3.5 Flash, intended for quick, high-volume subagent tasks rather than lengthy deliberation.

Best for Gemini 3.5 Flash-Lite:

  • high-throughput document parsing
  • low-cost translation
  • subagent classification

Borg lane ids: antigravity/gemini-3.5-flash-lite, google/gemini-3.5-flash-lite. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3.6 Flash

Gemini 3.6 Flash is Google's later third-generation workhorse following 3.5 Flash.

Best for Gemini 3.6 Flash:

  • rapid agentic coding loops
  • spatial reasoning from visual inputs
  • multistep tool execution

Borg lane ids: antigravity/gemini-3.6-flash-high, antigravity/gemini-3.6-flash-low, antigravity/gemini-3.6-flash-medium, antigravity/gemini-3.6-flash-tiered, google/gemini-3.6-flash. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3.7 Flash

Gemini 3.7 Flash follows 3.6 Flash in Google's third-generation workhorse line.

Best for Gemini 3.7 Flash:

  • video-informed agent workflows
  • software debugging with tools
  • multi-step coding at selected effort

Borg lane ids: antigravity/gemini-3.7-flash-high, antigravity/gemini-3.7-flash-low, antigravity/gemini-3.7-flash-medium, antigravity/gemini-3.7-flash-tiered, google/gemini-3.7-flash. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

Gemini 3.8 Flash

Gemini 3.8 Flash is Google's newer Flash workhorse, building on 3.7 Flash with a focus on long-horizon software engineering and autonomous, multi-step knowledge workflows.

Best for Gemini 3.8 Flash:

  • long-horizon coding agents
  • enterprise document workflows
  • autonomous multi-step tool use

Borg lane ids: antigravity/gemini-3.8-flash-high, antigravity/gemini-3.8-flash-low, antigravity/gemini-3.8-flash-medium, antigravity/gemini-3.8-flash-tiered. Open the model page for its full description, available specifications, vendor price sources, and routing roles.

How Fusion uses Google

Borg Fusion can route between models in the subscription catalog according to the task and the selected profile. Having a Google lane in the catalog does not mean that every Fusion profile will choose it for every request. No specific routing assignments for Google are published in this catalog. The model pages distinguish published routing roles from use cases, so you can compare options without assuming a benchmark result or a routing guarantee. You can also choose a listed lane directly if you want to control which model handles a request.

FAQ

Which Google models are in Borg?

The table and model details above list the current live Google models in Borg. Each linked page includes the model's lane ids and available specifications. Models awaiting identity review are not shown as live.

How does Fusion use Google?

Fusion profiles select from available lanes depending on the task. Consult each Google model page for its published Fusion roles; a missing role is not evidence that the model cannot be selected directly.

Can I select a Google lane directly?

Yes. Open a Google model page to find its Borg lane ids and the compatible gateway model field. Use the lane id rather than the public page slug when choosing a model directly.

Other providers

Compare Google with other makers in the same Borg subscription: Anthropic · OpenAI · xAI · Moonshot AI · Z.ai · Meta · Borg

All models