Modelflare
A VS Code extension that aggregates text models from multiple vendors into the VS Code Chat model selector through a Smart API Key. It supports BYOK workflows, streaming output, and tool calling.
Timeline (last 90 days)
The Modelflare VS Code extension has the ID modelflare.modelflare-vscode and requires VS Code 1.104 or newer.
Version 0.1.0 of the Modelflare VS Code extension used the extension ID modelflare.modelflare, and from version 0.1.1 onward it changed to modelflare.modelflare-vscode, so users need to re-enter their API Key after updating.
The Smart API Key is bound to the auto group, and on each request the server dynamically selects a candidate group among multiple routing groups according to the key policy, moving on to the next one when the channels of a group are exhausted.
The Smart API Key supports three routing strategies: Balanced (default, starting from the middle of the price range), Stability first (highest group multiplier first), and Low price first (lowest group multiplier first).
The Modelflare VS Code extension has only one setting, modelflare.baseUrl, whose default value global points to https://modelflare.dev/v1 and china points to https://origin.modelflare.dev/v1.
The Modelflare VS Code extension's API key is stored in VS Code secret storage and does not appear in settings.json.
Using Modelflare's BYOK models, the VS Code Chat panel can be driven without a Copilot subscription.
As of the article's publication, Modelflare's live model catalog contains over a dozen text models, covering families such as OpenAI GPT (e.g. gpt-6.1-sol, gpt-6-sol, gpt-5.5), Anthropic Claude (e.g. claude-opus-5-5, claude-sonnet-5-5, claude-haiku-5-5), Google Gemini (e.g. gemini-3.8-flash, gemini-3.5-flash, gemini-3.1-pro-preview), xAI Grok (grok-4.7), and DeepSeek.
The Modelflare VS Code extension is also published on Open VSX for use by compatible builds such as VSCodium.
All Facts (9)
The Modelflare VS Code extension has the ID modelflare.modelflare-vscode and requires VS Code 1.104 or newer.
50%UnverifiedVersion 0.1.0 of the Modelflare VS Code extension used the extension ID modelflare.modelflare, and from version 0.1.1 onward it changed to modelflare.modelflare-vscode, so users need to re-enter their API Key after updating.
50%UnverifiedThe Smart API Key is bound to the auto group, and on each request the server dynamically selects a candidate group among multiple routing groups according to the key policy, moving on to the next one when the channels of a group are exhausted.
50%UnverifiedThe Smart API Key supports three routing strategies: Balanced (default, starting from the middle of the price range), Stability first (highest group multiplier first), and Low price first (lowest group multiplier first).
50%UnverifiedThe Modelflare VS Code extension has only one setting, modelflare.baseUrl, whose default value global points to https://modelflare.dev/v1 and china points to https://origin.modelflare.dev/v1.
50%UnverifiedThe Modelflare VS Code extension's API key is stored in VS Code secret storage and does not appear in settings.json.
50%UnverifiedUsing Modelflare's BYOK models, the VS Code Chat panel can be driven without a Copilot subscription.
50%UnverifiedAs of the article's publication, Modelflare's live model catalog contains over a dozen text models, covering families such as OpenAI GPT (e.g. gpt-6.1-sol, gpt-6-sol, gpt-5.5), Anthropic Claude (e.g. claude-opus-5-5, claude-sonnet-5-5, claude-haiku-5-5), Google Gemini (e.g. gemini-3.8-flash, gemini-3.5-flash, gemini-3.1-pro-preview), xAI Grok (grok-4.7), and DeepSeek.
50%UnverifiedThe Modelflare VS Code extension is also published on Open VSX for use by compatible builds such as VSCodium.
50%