From 0e644f102ff0b9e74213f502c9c41cda7e7776e5 Mon Sep 17 00:00:00 2001 From: cte Date: Mon, 24 Feb 2025 12:48:55 -0800 Subject: [PATCH] Default to Claude 3.7 where appropriate --- .github/ISSUE_TEMPLATE/bug_report.yml | 2 +- src/core/Cline.ts | 2 +- src/core/webview/ClineProvider.ts | 50 +++---------------- src/shared/api.ts | 24 +++++---- src/test/suite/index.ts | 2 +- .../components/settings/GlamaModelPicker.tsx | 2 +- .../settings/OpenRouterModelPicker.tsx | 2 +- .../settings/RequestyModelPicker.tsx | 2 +- .../src/components/settings/SettingsView.tsx | 2 +- 9 files changed, 26 insertions(+), 62 deletions(-) diff --git a/.github/ISSUE_TEMPLATE/bug_report.yml b/.github/ISSUE_TEMPLATE/bug_report.yml index 501180c3d5..dc66b4f390 100644 --- a/.github/ISSUE_TEMPLATE/bug_report.yml +++ b/.github/ISSUE_TEMPLATE/bug_report.yml @@ -33,7 +33,7 @@ body: id: model attributes: label: Which Model are you using? - description: Please specify the model you're using (e.g. Claude 3.5 Sonnet) + description: Please specify the model you're using (e.g. Claude 3.7 Sonnet) validations: required: true - type: textarea diff --git a/src/core/Cline.ts b/src/core/Cline.ts index 12cf062406..f1f5e41b33 100644 --- a/src/core/Cline.ts +++ b/src/core/Cline.ts @@ -2792,7 +2792,7 @@ export class Cline { "mistake_limit_reached", this.api.getModel().id.includes("claude") ? `This may indicate a failure in his thought process or inability to use a tool properly, which can be mitigated with some user guidance (e.g. "Try breaking down the task into smaller steps").` - : "Roo Code uses complex prompts and iterative task execution that may be challenging for less capable models. For best results, it's recommended to use Claude 3.5 Sonnet for its advanced agentic coding capabilities.", + : "Roo Code uses complex prompts and iterative task execution that may be challenging for less capable models. For best results, it's recommended to use Claude 3.7 Sonnet for its advanced agentic coding capabilities.", ) if (response === "messageResponse") { userContent.push( diff --git a/src/core/webview/ClineProvider.ts b/src/core/webview/ClineProvider.ts index 6790224eca..b4819d9683 100644 --- a/src/core/webview/ClineProvider.ts +++ b/src/core/webview/ClineProvider.ts @@ -1900,23 +1900,7 @@ export class ClineProvider implements vscode.WebviewViewProvider { } const response = await axios.get("https://router.requesty.ai/v1/models", config) - /* - { - "id": "anthropic/claude-3-5-sonnet-20240620", - "object": "model", - "created": 1738243330, - "owned_by": "system", - "input_price": 0.000003, - "caching_price": 0.00000375, - "cached_price": 3E-7, - "output_price": 0.000015, - "max_output_tokens": 8192, - "context_window": 200000, - "supports_caching": true, - "description": "Anthropic's most intelligent model. Highest level of intelligence and capability" - }, - } - */ + if (response.data) { const rawModels = response.data.data const parsePrice = (price: any) => { @@ -2116,34 +2100,10 @@ export class ClineProvider implements vscode.WebviewViewProvider { ) const models: Record = {} + try { const response = await axios.get("https://openrouter.ai/api/v1/models") - /* - { - "id": "anthropic/claude-3.5-sonnet", - "name": "Anthropic: Claude 3.5 Sonnet", - "created": 1718841600, - "description": "Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: Autonomously writes, edits, and runs code with reasoning and troubleshooting\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal", - "context_length": 200000, - "architecture": { - "modality": "text+image-\u003Etext", - "tokenizer": "Claude", - "instruct_type": null - }, - "pricing": { - "prompt": "0.000003", - "completion": "0.000015", - "image": "0.0048", - "request": "0" - }, - "top_provider": { - "context_length": 200000, - "max_completion_tokens": 8192, - "is_moderated": true - }, - "per_request_limits": null - }, - */ + if (response.data?.data) { const rawModels = response.data.data const parsePrice = (price: any) => { @@ -2152,6 +2112,7 @@ export class ClineProvider implements vscode.WebviewViewProvider { } return undefined } + for (const rawModel of rawModels) { const modelInfo: ModelInfo = { maxTokens: rawModel.top_provider?.max_completion_tokens, @@ -2164,9 +2125,10 @@ export class ClineProvider implements vscode.WebviewViewProvider { } switch (rawModel.id) { + case "anthropic/claude-3.7-sonnet": case "anthropic/claude-3.5-sonnet": case "anthropic/claude-3.5-sonnet:beta": - // NOTE: this needs to be synced with api.ts/openrouter default model info + // NOTE: this needs to be synced with api.ts/openrouter default model info. modelInfo.supportsComputerUse = true modelInfo.supportsPromptCache = true modelInfo.cacheWritesPrice = 3.75 diff --git a/src/shared/api.ts b/src/shared/api.ts index 23fe60696c..4619d2930d 100644 --- a/src/shared/api.ts +++ b/src/shared/api.ts @@ -93,7 +93,7 @@ export interface ModelInfo { // Anthropic // https://docs.anthropic.com/en/docs/about-claude/models export type AnthropicModelId = keyof typeof anthropicModels -export const anthropicDefaultModelId: AnthropicModelId = "claude-3-5-sonnet-20241022" +export const anthropicDefaultModelId: AnthropicModelId = "claude-3-7-sonnet-20250219" export const anthropicModels = { "claude-3-7-sonnet-20250219": { maxTokens: 64_000, @@ -355,9 +355,9 @@ export const bedrockModels = { // Glama // https://glama.ai/models -export const glamaDefaultModelId = "anthropic/claude-3-5-sonnet" +export const glamaDefaultModelId = "anthropic/claude-3-7-sonnet" export const glamaDefaultModelInfo: ModelInfo = { - maxTokens: 8192, + maxTokens: 64_000, contextWindow: 200_000, supportsImages: true, supportsComputerUse: true, @@ -367,11 +367,14 @@ export const glamaDefaultModelInfo: ModelInfo = { cacheWritesPrice: 3.75, cacheReadsPrice: 0.3, description: - "The new Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: New Sonnet scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/anthropic/claude-3.5-sonnet) variant._", + "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. Claude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks. Read more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)", } +// Requesty +// https://requesty.ai/router-2 +export const requestyDefaultModelId = "anthropic/claude-3-7-sonnet-latest" export const requestyDefaultModelInfo: ModelInfo = { - maxTokens: 8192, + maxTokens: 64_000, contextWindow: 200_000, supportsImages: true, supportsComputerUse: true, @@ -381,15 +384,14 @@ export const requestyDefaultModelInfo: ModelInfo = { cacheWritesPrice: 3.75, cacheReadsPrice: 0.3, description: - "The new Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: New Sonnet scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/anthropic/claude-3.5-sonnet) variant._", + "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. Claude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks. Read more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)", } -export const requestyDefaultModelId = "anthropic/claude-3-5-sonnet" // OpenRouter // https://openrouter.ai/models?order=newest&supported_parameters=tools -export const openRouterDefaultModelId = "anthropic/claude-3.5-sonnet:beta" // will always exist in openRouterModels +export const openRouterDefaultModelId = "anthropic/claude-3.7-sonnet" export const openRouterDefaultModelInfo: ModelInfo = { - maxTokens: 8192, + maxTokens: 64_000, contextWindow: 200_000, supportsImages: true, supportsComputerUse: true, @@ -399,13 +401,13 @@ export const openRouterDefaultModelInfo: ModelInfo = { cacheWritesPrice: 3.75, cacheReadsPrice: 0.3, description: - "The new Claude 3.5 Sonnet delivers better-than-Opus capabilities, faster-than-Sonnet speeds, at the same Sonnet prices. Sonnet is particularly good at:\n\n- Coding: New Sonnet scores ~49% on SWE-Bench Verified, higher than the last best score, and without any fancy prompt scaffolding\n- Data science: Augments human data science expertise; navigates unstructured data while using multiple tools for insights\n- Visual processing: excelling at interpreting charts, graphs, and images, accurately transcribing text to derive insights beyond just the text alone\n- Agentic tasks: exceptional tool use, making it great at agentic tasks (i.e. complex, multi-step problem solving tasks that require engaging with other systems)\n\n#multimodal\n\n_This is a faster endpoint, made available in collaboration with Anthropic, that is self-moderated: response moderation happens on the provider's side instead of OpenRouter's. For requests that pass moderation, it's identical to the [Standard](/anthropic/claude-3.5-sonnet) variant._", + "Claude 3.7 Sonnet is an advanced large language model with improved reasoning, coding, and problem-solving capabilities. It introduces a hybrid reasoning approach, allowing users to choose between rapid responses and extended, step-by-step processing for complex tasks. The model demonstrates notable improvements in coding, particularly in front-end development and full-stack updates, and excels in agentic workflows, where it can autonomously navigate multi-step processes. Claude 3.7 Sonnet maintains performance parity with its predecessor in standard mode while offering an extended reasoning mode for enhanced accuracy in math, coding, and instruction-following tasks. Read more at the [blog post here](https://www.anthropic.com/news/claude-3-7-sonnet)", } // Vertex AI // https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-claude export type VertexModelId = keyof typeof vertexModels -export const vertexDefaultModelId: VertexModelId = "claude-3-5-sonnet-v2@20241022" +export const vertexDefaultModelId: VertexModelId = "claude-3-7-sonnet@20250219" export const vertexModels = { "claude-3-7-sonnet@20250219": { maxTokens: 8192, diff --git a/src/test/suite/index.ts b/src/test/suite/index.ts index ffb8de7473..540be7cef8 100644 --- a/src/test/suite/index.ts +++ b/src/test/suite/index.ts @@ -39,7 +39,7 @@ export async function run(): Promise { : await globalThis.extension.activate() globalThis.provider = globalThis.api.sidebarProvider await globalThis.provider.updateGlobalState("apiProvider", "openrouter") - await globalThis.provider.updateGlobalState("openRouterModelId", "anthropic/claude-3.5-sonnet") + await globalThis.provider.updateGlobalState("openRouterModelId", "anthropic/claude-3.7-sonnet") await globalThis.provider.storeSecret( "openRouterApiKey", process.env.OPENROUTER_API_KEY || "sk-or-v1-fake-api-key", diff --git a/webview-ui/src/components/settings/GlamaModelPicker.tsx b/webview-ui/src/components/settings/GlamaModelPicker.tsx index cb813a0d05..37e326d8f8 100644 --- a/webview-ui/src/components/settings/GlamaModelPicker.tsx +++ b/webview-ui/src/components/settings/GlamaModelPicker.tsx @@ -10,6 +10,6 @@ export const GlamaModelPicker = () => ( refreshMessageType="refreshGlamaModels" serviceName="Glama" serviceUrl="https://glama.ai/models" - recommendedModel="anthropic/claude-3-5-sonnet" + recommendedModel="anthropic/claude-3-7-sonnet" /> ) diff --git a/webview-ui/src/components/settings/OpenRouterModelPicker.tsx b/webview-ui/src/components/settings/OpenRouterModelPicker.tsx index 9111407cd6..c773478e54 100644 --- a/webview-ui/src/components/settings/OpenRouterModelPicker.tsx +++ b/webview-ui/src/components/settings/OpenRouterModelPicker.tsx @@ -10,6 +10,6 @@ export const OpenRouterModelPicker = () => ( refreshMessageType="refreshOpenRouterModels" serviceName="OpenRouter" serviceUrl="https://openrouter.ai/models" - recommendedModel="anthropic/claude-3.5-sonnet:beta" + recommendedModel="anthropic/claude-3.7-sonnet" /> ) diff --git a/webview-ui/src/components/settings/RequestyModelPicker.tsx b/webview-ui/src/components/settings/RequestyModelPicker.tsx index e0759a43ba..c65067068a 100644 --- a/webview-ui/src/components/settings/RequestyModelPicker.tsx +++ b/webview-ui/src/components/settings/RequestyModelPicker.tsx @@ -16,7 +16,7 @@ export const RequestyModelPicker = () => { }} serviceName="Requesty" serviceUrl="https://requesty.ai" - recommendedModel="anthropic/claude-3-5-sonnet-latest" + recommendedModel="anthropic/claude-3-7-sonnet-latest" /> ) } diff --git a/webview-ui/src/components/settings/SettingsView.tsx b/webview-ui/src/components/settings/SettingsView.tsx index 495bf49bd7..0d80580b49 100644 --- a/webview-ui/src/components/settings/SettingsView.tsx +++ b/webview-ui/src/components/settings/SettingsView.tsx @@ -765,7 +765,7 @@ const SettingsView = forwardRef(({ onDone }, color: "var(--vscode-descriptionForeground)", }}> When enabled, Roo will be able to edit files more quickly and will automatically reject - truncated full-file writes. Works best with the latest Claude 3.5 Sonnet model. + truncated full-file writes. Works best with the latest Claude 3.7 Sonnet model.

{diffEnabled && (