feat(ai): add Fireworks provider support closes #3519

2026-04-22 01:08:57 +02:00
parent 92b08abd3b
commit 0bb0a58466
12 changed files with 400 additions and 0 deletions
--- a/packages/ai/CHANGELOG.md
+++ b/packages/ai/CHANGELOG.md
@@ -2,6 +2,10 @@

 ## [Unreleased]

+### Added
+
+- Added Fireworks provider support via Fireworks' Anthropic-compatible Messages API, including built-in models sourced from models.dev and `FIREWORKS_API_KEY` auth ([#3519](https://github.com/badlogic/pi-mono/issues/3519))
+
 ### Fixed

 - Hardened Anthropic streaming against malformed tool-call JSON by owning SSE parsing with defensive JSON repair, replacing the deprecated `fine-grained-tool-streaming` beta header with per-tool `eager_input_streaming`, and updating stale test model references ([#3175](https://github.com/badlogic/pi-mono/issues/3175))
--- a/packages/ai/README.md
+++ b/packages/ai/README.md
@@ -66,6 +66,7 @@ Unified LLM API with automatic model discovery, provider configuration, token an
 - **Amazon Bedrock**
 - **OpenCode Zen**
 - **OpenCode Go**
+- **Fireworks** (uses Anthropic-compatible API)
 - **Kimi For Coding** (Moonshot AI, uses Anthropic-compatible API)
 - **Any OpenAI-compatible API**: Ollama, vLLM, LM Studio, etc.

@@ -1025,6 +1026,7 @@ In Node.js environments, you can set environment variables to avoid passing API
 | Groq | `GROQ_API_KEY` |
 | Cerebras | `CEREBRAS_API_KEY` |
 | xAI | `XAI_API_KEY` |
+| Fireworks | `FIREWORKS_API_KEY` |
 | OpenRouter | `OPENROUTER_API_KEY` |
 | Vercel AI Gateway | `AI_GATEWAY_API_KEY` |
 | zAI | `ZAI_API_KEY` |
--- a/packages/ai/scripts/generate-models.ts
+++ b/packages/ai/scripts/generate-models.ts
@@ -468,6 +468,33 @@ async function loadModelsDevData(): Promise<Model<any>[]> {
 			}
 		}

+		// Process Fireworks models
+		if (data["fireworks-ai"]?.models) {
+			for (const [modelId, model] of Object.entries(data["fireworks-ai"].models)) {
+				const m = model as ModelsDevModel;
+				if (m.tool_call !== true) continue;
+
+				models.push({
+					id: modelId,
+					name: m.name || modelId,
+					api: "anthropic-messages",
+					provider: "fireworks",
+					// Fireworks Anthropic-compatible API - SDK appends /v1/messages
+					baseUrl: "https://api.fireworks.ai/inference",
+					reasoning: m.reasoning === true,
+					input: m.modalities?.input?.includes("image") ? ["text", "image"] : ["text"],
+					cost: {
+						input: m.cost?.input || 0,
+						output: m.cost?.output || 0,
+						cacheRead: m.cost?.cache_read || 0,
+						cacheWrite: m.cost?.cache_write || 0,
+					},
+					contextWindow: m.limit?.context || 4096,
+					maxTokens: m.limit?.output || 4096,
+				});
+			}
+		}
+
 		// Process OpenCode models (Zen and Go)
 		// API mapping based on provider.npm field:
 		// - @ai-sdk/openai → openai-responses
--- a/packages/ai/src/env-api-keys.ts
+++ b/packages/ai/src/env-api-keys.ts
@@ -123,6 +123,7 @@ export function getEnvApiKey(provider: any): string | undefined {
 		minimax: "MINIMAX_API_KEY",
 		"minimax-cn": "MINIMAX_CN_API_KEY",
 		huggingface: "HF_TOKEN",
+		fireworks: "FIREWORKS_API_KEY",
 		opencode: "OPENCODE_API_KEY",
 		"opencode-go": "OPENCODE_API_KEY",
 		"kimi-coding": "KIMI_API_KEY",
--- a/packages/ai/src/models.generated.ts
+++ b/packages/ai/src/models.generated.ts
@@ -2698,6 +2698,314 @@ export const MODELS = {
 			maxTokens: 40000,
 		} satisfies Model<"openai-completions">,
 	},
+	"fireworks": {
+		"accounts/fireworks/models/deepseek-v3p1": {
+			id: "accounts/fireworks/models/deepseek-v3p1",
+			name: "DeepSeek V3.1",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.56,
+				output: 1.68,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 163840,
+			maxTokens: 163840,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/deepseek-v3p2": {
+			id: "accounts/fireworks/models/deepseek-v3p2",
+			name: "DeepSeek V3.2",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.56,
+				output: 1.68,
+				cacheRead: 0.28,
+				cacheWrite: 0,
+			},
+			contextWindow: 160000,
+			maxTokens: 160000,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/glm-4p5": {
+			id: "accounts/fireworks/models/glm-4p5",
+			name: "GLM 4.5",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.55,
+				output: 2.19,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 131072,
+			maxTokens: 131072,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/glm-4p5-air": {
+			id: "accounts/fireworks/models/glm-4p5-air",
+			name: "GLM 4.5 Air",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.22,
+				output: 0.88,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 131072,
+			maxTokens: 131072,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/glm-4p7": {
+			id: "accounts/fireworks/models/glm-4p7",
+			name: "GLM 4.7",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.6,
+				output: 2.2,
+				cacheRead: 0.3,
+				cacheWrite: 0,
+			},
+			contextWindow: 198000,
+			maxTokens: 198000,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/glm-5": {
+			id: "accounts/fireworks/models/glm-5",
+			name: "GLM 5",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 1,
+				output: 3.2,
+				cacheRead: 0.5,
+				cacheWrite: 0,
+			},
+			contextWindow: 202752,
+			maxTokens: 131072,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/glm-5p1": {
+			id: "accounts/fireworks/models/glm-5p1",
+			name: "GLM 5.1",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 1.4,
+				output: 4.4,
+				cacheRead: 0.26,
+				cacheWrite: 0,
+			},
+			contextWindow: 202800,
+			maxTokens: 131072,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/gpt-oss-120b": {
+			id: "accounts/fireworks/models/gpt-oss-120b",
+			name: "GPT OSS 120B",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.15,
+				output: 0.6,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 131072,
+			maxTokens: 32768,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/gpt-oss-20b": {
+			id: "accounts/fireworks/models/gpt-oss-20b",
+			name: "GPT OSS 20B",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.05,
+				output: 0.2,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 131072,
+			maxTokens: 32768,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/kimi-k2-instruct": {
+			id: "accounts/fireworks/models/kimi-k2-instruct",
+			name: "Kimi K2 Instruct",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: false,
+			input: ["text"],
+			cost: {
+				input: 1,
+				output: 3,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 128000,
+			maxTokens: 16384,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/kimi-k2-thinking": {
+			id: "accounts/fireworks/models/kimi-k2-thinking",
+			name: "Kimi K2 Thinking",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.6,
+				output: 2.5,
+				cacheRead: 0.3,
+				cacheWrite: 0,
+			},
+			contextWindow: 256000,
+			maxTokens: 256000,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/kimi-k2p5": {
+			id: "accounts/fireworks/models/kimi-k2p5",
+			name: "Kimi K2.5",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 0.6,
+				output: 3,
+				cacheRead: 0.1,
+				cacheWrite: 0,
+			},
+			contextWindow: 256000,
+			maxTokens: 256000,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/kimi-k2p6": {
+			id: "accounts/fireworks/models/kimi-k2p6",
+			name: "Kimi K2.6",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 0.95,
+				output: 4,
+				cacheRead: 0.16,
+				cacheWrite: 0,
+			},
+			contextWindow: 262000,
+			maxTokens: 262000,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/minimax-m2p1": {
+			id: "accounts/fireworks/models/minimax-m2p1",
+			name: "MiniMax-M2.1",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.3,
+				output: 1.2,
+				cacheRead: 0.03,
+				cacheWrite: 0,
+			},
+			contextWindow: 200000,
+			maxTokens: 200000,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/minimax-m2p5": {
+			id: "accounts/fireworks/models/minimax-m2p5",
+			name: "MiniMax-M2.5",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.3,
+				output: 1.2,
+				cacheRead: 0.03,
+				cacheWrite: 0,
+			},
+			contextWindow: 196608,
+			maxTokens: 196608,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/minimax-m2p7": {
+			id: "accounts/fireworks/models/minimax-m2p7",
+			name: "MiniMax-M2.7",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text"],
+			cost: {
+				input: 0.3,
+				output: 1.2,
+				cacheRead: 0.03,
+				cacheWrite: 0,
+			},
+			contextWindow: 196608,
+			maxTokens: 196608,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/models/qwen3p6-plus": {
+			id: "accounts/fireworks/models/qwen3p6-plus",
+			name: "Qwen 3.6 Plus",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 0.5,
+				output: 3,
+				cacheRead: 0.1,
+				cacheWrite: 0,
+			},
+			contextWindow: 128000,
+			maxTokens: 8192,
+		} satisfies Model<"anthropic-messages">,
+		"accounts/fireworks/routers/kimi-k2p5-turbo": {
+			id: "accounts/fireworks/routers/kimi-k2p5-turbo",
+			name: "Kimi K2.5 Turbo (firepass)",
+			api: "anthropic-messages",
+			provider: "fireworks",
+			baseUrl: "https://api.fireworks.ai/inference",
+			reasoning: true,
+			input: ["text", "image"],
+			cost: {
+				input: 0,
+				output: 0,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+			contextWindow: 256000,
+			maxTokens: 256000,
+		} satisfies Model<"anthropic-messages">,
+	},
 	"github-copilot": {
 		"claude-haiku-4.5": {
 			id: "claude-haiku-4.5",
--- a/packages/ai/src/types.ts
+++ b/packages/ai/src/types.ts
@@ -37,6 +37,7 @@ export type KnownProvider =
 	| "minimax"
 	| "minimax-cn"
 	| "huggingface"
+	| "fireworks"
 	| "opencode"
 	| "opencode-go"
 	| "kimi-coding";
--- a/packages/ai/test/fireworks-models.test.ts
+++ b/packages/ai/test/fireworks-models.test.ts
@@ -0,0 +1,49 @@
+import { afterEach, describe, expect, it } from "vitest";
+import { getEnvApiKey } from "../src/env-api-keys.js";
+import { getModel } from "../src/models.js";
+
+const originalFireworksApiKey = process.env.FIREWORKS_API_KEY;
+
+afterEach(() => {
+	if (originalFireworksApiKey === undefined) {
+		delete process.env.FIREWORKS_API_KEY;
+	} else {
+		process.env.FIREWORKS_API_KEY = originalFireworksApiKey;
+	}
+});
+
+describe("Fireworks models", () => {
+	it("registers the default Kimi K2.6 model via Anthropic-compatible Messages API", () => {
+		const model = getModel("fireworks", "accounts/fireworks/models/kimi-k2p6");
+
+		expect(model).toBeDefined();
+		expect(model.api).toBe("anthropic-messages");
+		expect(model.provider).toBe("fireworks");
+		expect(model.baseUrl).toBe("https://api.fireworks.ai/inference");
+		expect(model.reasoning).toBe(true);
+		expect(model.input).toEqual(["text", "image"]);
+		expect(model.contextWindow).toBe(262000);
+		expect(model.maxTokens).toBe(262000);
+		expect(model.cost).toEqual({
+			input: 0.95,
+			output: 4,
+			cacheRead: 0.16,
+			cacheWrite: 0,
+		});
+	});
+
+	it("registers the Fire Pass turbo router model", () => {
+		const model = getModel("fireworks", "accounts/fireworks/routers/kimi-k2p5-turbo");
+
+		expect(model).toBeDefined();
+		expect(model.api).toBe("anthropic-messages");
+		expect(model.baseUrl).toBe("https://api.fireworks.ai/inference");
+		expect(model.input).toEqual(["text", "image"]);
+	});
+
+	it("resolves FIREWORKS_API_KEY from the environment", () => {
+		process.env.FIREWORKS_API_KEY = "test-fireworks-key";
+
+		expect(getEnvApiKey("fireworks")).toBe("test-fireworks-key");
+	});
+});
--- a/packages/coding-agent/CHANGELOG.md
+++ b/packages/coding-agent/CHANGELOG.md
@@ -2,6 +2,10 @@

 ## [Unreleased]

+### Added
+
+- Added built-in Fireworks provider support, including `FIREWORKS_API_KEY` setup/docs and the default Fireworks model `accounts/fireworks/models/kimi-k2p6` ([#3519](https://github.com/badlogic/pi-mono/issues/3519))
+
 ### Fixed

 - Added configurable interactive inline image width via `terminal.imageWidthCells` and `/settings`, so tool-output images are no longer hard-capped to 60 terminal cells ([#3508](https://github.com/badlogic/pi-mono/issues/3508))
--- a/packages/coding-agent/README.md
+++ b/packages/coding-agent/README.md
@@ -120,6 +120,7 @@ For each built-in provider, pi maintains a list of tool-capable models, updated
 - OpenCode Zen
 - OpenCode Go
 - Hugging Face
+- Fireworks
 - Kimi For Coding
 - MiniMax

--- a/packages/coding-agent/docs/providers.md
+++ b/packages/coding-agent/docs/providers.md
@@ -67,6 +67,7 @@ pi
 | OpenCode Zen | `OPENCODE_API_KEY` | `opencode` |
 | OpenCode Go | `OPENCODE_API_KEY` | `opencode-go` |
 | Hugging Face | `HF_TOKEN` | `huggingface` |
+| Fireworks | `FIREWORKS_API_KEY` | `fireworks` |
 | Kimi For Coding | `KIMI_API_KEY` | `kimi-coding` |
 | MiniMax | `MINIMAX_API_KEY` | `minimax` |
 | MiniMax (China) | `MINIMAX_CN_API_KEY` | `minimax-cn` |
--- a/packages/coding-agent/src/cli/args.ts
+++ b/packages/coding-agent/src/cli/args.ts
@@ -303,6 +303,7 @@ ${chalk.bold("Environment Variables:")}
  GROQ_API_KEY                     - Groq API key
  CEREBRAS_API_KEY                 - Cerebras API key
  XAI_API_KEY                      - xAI Grok API key
+  FIREWORKS_API_KEY                - Fireworks API key
  OPENROUTER_API_KEY               - OpenRouter API key
  AI_GATEWAY_API_KEY               - Vercel AI Gateway API key
  ZAI_API_KEY                      - ZAI API key
--- a/packages/coding-agent/src/core/model-resolver.ts
+++ b/packages/coding-agent/src/core/model-resolver.ts
@@ -32,6 +32,7 @@ export const defaultModelPerProvider: Record<KnownProvider, string> = {
 	minimax: "MiniMax-M2.7",
 	"minimax-cn": "MiniMax-M2.7",
 	huggingface: "moonshotai/Kimi-K2.5",
+	fireworks: "accounts/fireworks/models/kimi-k2p6",
 	opencode: "claude-opus-4-6",
 	"opencode-go": "kimi-k2.5",
 	"kimi-coding": "kimi-for-coding",