mirror of
https://github.com/getcompanion-ai/co-mono.git
synced 2026-04-15 08:03:39 +00:00
docs(ai): Improve README with model discovery and capabilities documentation
- Added note that library only includes tool-calling capable models - Added Model Discovery section showing how to enumerate models - Added examples for finding models with specific capabilities - Added cache read/write costs to model capabilities display - Clarified that models are auto-fetched from APIs at build time
This commit is contained in:
Mario Zechner
parent
d46a98ec10
commit
dae40167a3
10 changed files with 167 additions and 108 deletions
22
package-lock.json
generated
22
package-lock.json
generated
|
|
@ -661,10 +661,6 @@
|
|||
"dev": true,
|
||||
"license": "MIT"
|
||||
},
|
||||
"node_modules/@mariozechner/ai": {
|
||||
"resolved": "packages/ai",
|
||||
"link": true
|
||||
},
|
||||
"node_modules/@mariozechner/pi": {
|
||||
"resolved": "packages/pods",
|
||||
"link": true
|
||||
|
|
@ -673,6 +669,10 @@
|
|||
"resolved": "packages/agent",
|
||||
"link": true
|
||||
},
|
||||
"node_modules/@mariozechner/pi-ai": {
|
||||
"resolved": "packages/ai",
|
||||
"link": true
|
||||
},
|
||||
"node_modules/@mariozechner/pi-tui": {
|
||||
"resolved": "packages/tui",
|
||||
"link": true
|
||||
|
|
@ -2716,10 +2716,10 @@
|
|||
},
|
||||
"packages/agent": {
|
||||
"name": "@mariozechner/pi-agent",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"license": "MIT",
|
||||
"dependencies": {
|
||||
"@mariozechner/pi-tui": "^0.5.8",
|
||||
"@mariozechner/pi-tui": "^0.5.9",
|
||||
"@types/glob": "^8.1.0",
|
||||
"chalk": "^5.5.0",
|
||||
"glob": "^11.0.3",
|
||||
|
|
@ -3097,8 +3097,8 @@
|
|||
}
|
||||
},
|
||||
"packages/ai": {
|
||||
"name": "@mariozechner/ai",
|
||||
"version": "0.5.9",
|
||||
"name": "@mariozechner/pi-ai",
|
||||
"version": "0.5.10",
|
||||
"license": "MIT",
|
||||
"dependencies": {
|
||||
"@anthropic-ai/sdk": "^0.60.0",
|
||||
|
|
@ -3134,10 +3134,10 @@
|
|||
},
|
||||
"packages/pods": {
|
||||
"name": "@mariozechner/pi",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"license": "MIT",
|
||||
"dependencies": {
|
||||
"@mariozechner/pi-agent": "^0.5.8",
|
||||
"@mariozechner/pi-agent": "^0.5.9",
|
||||
"chalk": "^5.5.0"
|
||||
},
|
||||
"bin": {
|
||||
|
|
@ -3150,7 +3150,7 @@
|
|||
},
|
||||
"packages/tui": {
|
||||
"name": "@mariozechner/pi-tui",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"license": "MIT",
|
||||
"dependencies": {
|
||||
"@types/mime-types": "^2.1.4",
|
||||
|
|
|
|||
4
packages/agent/package-lock.json
generated
4
packages/agent/package-lock.json
generated
|
|
@ -1,12 +1,12 @@
|
|||
{
|
||||
"name": "@mariozechner/pi-agent",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"lockfileVersion": 3,
|
||||
"requires": true,
|
||||
"packages": {
|
||||
"": {
|
||||
"name": "@mariozechner/pi-agent",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"license": "MIT",
|
||||
"dependencies": {
|
||||
"@mariozechner/tui": "^0.1.1",
|
||||
|
|
|
|||
|
|
@ -1,6 +1,6 @@
|
|||
{
|
||||
"name": "@mariozechner/pi-agent",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"description": "General-purpose agent with tool calling and session persistence",
|
||||
"type": "module",
|
||||
"bin": {
|
||||
|
|
@ -18,7 +18,7 @@
|
|||
"prepublishOnly": "npm run clean && npm run build"
|
||||
},
|
||||
"dependencies": {
|
||||
"@mariozechner/pi-tui": "^0.5.9",
|
||||
"@mariozechner/pi-tui": "^0.5.10",
|
||||
"@types/glob": "^8.1.0",
|
||||
"chalk": "^5.5.0",
|
||||
"glob": "^11.0.3",
|
||||
|
|
|
|||
|
|
@ -2,6 +2,8 @@
|
|||
|
||||
Unified LLM API with automatic model discovery, provider configuration, token and cost tracking, and simple context persistence and hand-off to other models mid-session.
|
||||
|
||||
**Note**: This library only includes models that support tool calling (function calling), as this is essential for agentic workflows.
|
||||
|
||||
## Supported Providers
|
||||
|
||||
- **OpenAI**
|
||||
|
|
@ -203,6 +205,63 @@ const model = {
|
|||
const llm = new OpenAICompletionsLLM(model, 'your-api-key');
|
||||
```
|
||||
|
||||
## Model Discovery
|
||||
|
||||
All models in this library support tool calling. Models are automatically fetched from OpenRouter and models.dev APIs at build time.
|
||||
|
||||
### List Available Models
|
||||
```typescript
|
||||
import { PROVIDERS } from '@mariozechner/pi-ai';
|
||||
|
||||
// List all OpenAI models (all support tool calling)
|
||||
for (const [modelId, model] of Object.entries(PROVIDERS.openai.models)) {
|
||||
console.log(`${modelId}: ${model.name}`);
|
||||
console.log(` Context: ${model.contextWindow} tokens`);
|
||||
console.log(` Reasoning: ${model.reasoning}`);
|
||||
console.log(` Vision: ${model.input.includes('image')}`);
|
||||
console.log(` Cost: $${model.cost.input}/$${model.cost.output} per million tokens`);
|
||||
}
|
||||
|
||||
// Find all models with reasoning support
|
||||
const reasoningModels = [];
|
||||
for (const provider of Object.values(PROVIDERS)) {
|
||||
for (const model of Object.values(provider.models)) {
|
||||
if (model.reasoning) {
|
||||
reasoningModels.push(model);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// Find all vision-capable models
|
||||
const visionModels = [];
|
||||
for (const provider of Object.values(PROVIDERS)) {
|
||||
for (const model of Object.values(provider.models)) {
|
||||
if (model.input.includes('image')) {
|
||||
visionModels.push(model);
|
||||
}
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### Check Model Capabilities
|
||||
```typescript
|
||||
import { getModel } from '@mariozechner/pi-ai';
|
||||
|
||||
const model = getModel('openai', 'gpt-4o-mini');
|
||||
if (model) {
|
||||
console.log(`Model: ${model.name}`);
|
||||
console.log(`Provider: ${model.provider}`);
|
||||
console.log(`Context window: ${model.contextWindow} tokens`);
|
||||
console.log(`Max output: ${model.maxTokens} tokens`);
|
||||
console.log(`Supports reasoning: ${model.reasoning}`);
|
||||
console.log(`Supports images: ${model.input.includes('image')}`);
|
||||
console.log(`Input cost: $${model.cost.input} per million tokens`);
|
||||
console.log(`Output cost: $${model.cost.output} per million tokens`);
|
||||
console.log(`Cache read cost: $${model.cost.cacheRead} per million tokens`);
|
||||
console.log(`Cache write cost: $${model.cost.cacheWrite} per million tokens`);
|
||||
}
|
||||
```
|
||||
|
||||
## Environment Variables
|
||||
|
||||
Set these environment variables to use `createLLM` without passing API keys:
|
||||
|
|
|
|||
|
|
@ -1,6 +1,6 @@
|
|||
{
|
||||
"name": "@mariozechner/pi-ai",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"description": "Unified LLM API with automatic model discovery and provider configuration",
|
||||
"type": "module",
|
||||
"main": "./dist/index.js",
|
||||
|
|
|
|||
|
|
@ -1598,22 +1598,6 @@ export const PROVIDERS = {
|
|||
contextWindow: 131072,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"meta-llama/llama-3.1-405b-instruct": {
|
||||
id: "meta-llama/llama-3.1-405b-instruct",
|
||||
name: "Meta: Llama 3.1 405B Instruct",
|
||||
provider: "openrouter",
|
||||
baseUrl: "https://openrouter.ai/api/v1",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.7999999999999999,
|
||||
output: 0.7999999999999999,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 32768,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"meta-llama/llama-3.1-70b-instruct": {
|
||||
id: "meta-llama/llama-3.1-70b-instruct",
|
||||
name: "Meta: Llama 3.1 70B Instruct",
|
||||
|
|
@ -1630,6 +1614,22 @@ export const PROVIDERS = {
|
|||
contextWindow: 131072,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"meta-llama/llama-3.1-405b-instruct": {
|
||||
id: "meta-llama/llama-3.1-405b-instruct",
|
||||
name: "Meta: Llama 3.1 405B Instruct",
|
||||
provider: "openrouter",
|
||||
baseUrl: "https://openrouter.ai/api/v1",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.7999999999999999,
|
||||
output: 0.7999999999999999,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 32768,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"mistralai/mistral-nemo": {
|
||||
id: "mistralai/mistral-nemo",
|
||||
name: "Mistral: Mistral Nemo",
|
||||
|
|
@ -1646,6 +1646,22 @@ export const PROVIDERS = {
|
|||
contextWindow: 32000,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"mistralai/mistral-7b-instruct-v0.3": {
|
||||
id: "mistralai/mistral-7b-instruct-v0.3",
|
||||
name: "Mistral: Mistral 7B Instruct v0.3",
|
||||
provider: "openrouter",
|
||||
baseUrl: "https://openrouter.ai/api/v1",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.028,
|
||||
output: 0.054,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 32768,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"mistralai/mistral-7b-instruct:free": {
|
||||
id: "mistralai/mistral-7b-instruct:free",
|
||||
name: "Mistral: Mistral 7B Instruct (free)",
|
||||
|
|
@ -1678,22 +1694,6 @@ export const PROVIDERS = {
|
|||
contextWindow: 32768,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"mistralai/mistral-7b-instruct-v0.3": {
|
||||
id: "mistralai/mistral-7b-instruct-v0.3",
|
||||
name: "Mistral: Mistral 7B Instruct v0.3",
|
||||
provider: "openrouter",
|
||||
baseUrl: "https://openrouter.ai/api/v1",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.028,
|
||||
output: 0.054,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 32768,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"microsoft/phi-3-mini-128k-instruct": {
|
||||
id: "microsoft/phi-3-mini-128k-instruct",
|
||||
name: "Microsoft: Phi-3 Mini 128K Instruct",
|
||||
|
|
@ -1726,22 +1726,6 @@ export const PROVIDERS = {
|
|||
contextWindow: 128000,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"meta-llama/llama-3-8b-instruct": {
|
||||
id: "meta-llama/llama-3-8b-instruct",
|
||||
name: "Meta: Llama 3 8B Instruct",
|
||||
provider: "openrouter",
|
||||
baseUrl: "https://openrouter.ai/api/v1",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.03,
|
||||
output: 0.06,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 8192,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"meta-llama/llama-3-70b-instruct": {
|
||||
id: "meta-llama/llama-3-70b-instruct",
|
||||
name: "Meta: Llama 3 70B Instruct",
|
||||
|
|
@ -1758,6 +1742,22 @@ export const PROVIDERS = {
|
|||
contextWindow: 8192,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"meta-llama/llama-3-8b-instruct": {
|
||||
id: "meta-llama/llama-3-8b-instruct",
|
||||
name: "Meta: Llama 3 8B Instruct",
|
||||
provider: "openrouter",
|
||||
baseUrl: "https://openrouter.ai/api/v1",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.03,
|
||||
output: 0.06,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 8192,
|
||||
maxTokens: 16384,
|
||||
} satisfies Model,
|
||||
"mistralai/mixtral-8x22b-instruct": {
|
||||
id: "mistralai/mixtral-8x22b-instruct",
|
||||
name: "Mistral: Mixtral 8x22B Instruct",
|
||||
|
|
@ -1854,22 +1854,6 @@ export const PROVIDERS = {
|
|||
contextWindow: 128000,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"mistralai/mistral-small": {
|
||||
id: "mistralai/mistral-small",
|
||||
name: "Mistral Small",
|
||||
provider: "openrouter",
|
||||
baseUrl: "https://openrouter.ai/api/v1",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.19999999999999998,
|
||||
output: 0.6,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 32768,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"mistralai/mistral-tiny": {
|
||||
id: "mistralai/mistral-tiny",
|
||||
name: "Mistral Tiny",
|
||||
|
|
@ -1886,6 +1870,22 @@ export const PROVIDERS = {
|
|||
contextWindow: 32768,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"mistralai/mistral-small": {
|
||||
id: "mistralai/mistral-small",
|
||||
name: "Mistral Small",
|
||||
provider: "openrouter",
|
||||
baseUrl: "https://openrouter.ai/api/v1",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.19999999999999998,
|
||||
output: 0.6,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 32768,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"mistralai/mixtral-8x7b-instruct": {
|
||||
id: "mistralai/mixtral-8x7b-instruct",
|
||||
name: "Mistral: Mixtral 8x7B Instruct",
|
||||
|
|
@ -2473,21 +2473,6 @@ export const PROVIDERS = {
|
|||
contextWindow: 16385,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"gpt-3.5-turbo": {
|
||||
id: "gpt-3.5-turbo",
|
||||
name: "OpenAI: GPT-3.5 Turbo",
|
||||
provider: "openai",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.5,
|
||||
output: 1.5,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 16385,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"gpt-4": {
|
||||
id: "gpt-4",
|
||||
name: "OpenAI: GPT-4",
|
||||
|
|
@ -2518,6 +2503,21 @@ export const PROVIDERS = {
|
|||
contextWindow: 8191,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
"gpt-3.5-turbo": {
|
||||
id: "gpt-3.5-turbo",
|
||||
name: "OpenAI: GPT-3.5 Turbo",
|
||||
provider: "openai",
|
||||
reasoning: false,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
input: 0.5,
|
||||
output: 1.5,
|
||||
cacheRead: 0,
|
||||
cacheWrite: 0,
|
||||
},
|
||||
contextWindow: 16385,
|
||||
maxTokens: 4096,
|
||||
} satisfies Model,
|
||||
},
|
||||
},
|
||||
anthropic: {
|
||||
|
|
@ -2597,9 +2597,9 @@ export const PROVIDERS = {
|
|||
contextWindow: 200000,
|
||||
maxTokens: 64000,
|
||||
} satisfies Model,
|
||||
"claude-3-5-haiku-latest": {
|
||||
id: "claude-3-5-haiku-latest",
|
||||
name: "Anthropic: Claude 3.5 Haiku",
|
||||
"claude-3-5-haiku-20241022": {
|
||||
id: "claude-3-5-haiku-20241022",
|
||||
name: "Anthropic: Claude 3.5 Haiku (2024-10-22)",
|
||||
provider: "anthropic",
|
||||
reasoning: false,
|
||||
input: ["text", "image"],
|
||||
|
|
@ -2612,9 +2612,9 @@ export const PROVIDERS = {
|
|||
contextWindow: 200000,
|
||||
maxTokens: 8192,
|
||||
} satisfies Model,
|
||||
"claude-3-5-haiku-20241022": {
|
||||
id: "claude-3-5-haiku-20241022",
|
||||
name: "Anthropic: Claude 3.5 Haiku (2024-10-22)",
|
||||
"claude-3-5-haiku-latest": {
|
||||
id: "claude-3-5-haiku-latest",
|
||||
name: "Anthropic: Claude 3.5 Haiku",
|
||||
provider: "anthropic",
|
||||
reasoning: false,
|
||||
input: ["text", "image"],
|
||||
|
|
|
|||
4
packages/pods/package-lock.json
generated
4
packages/pods/package-lock.json
generated
|
|
@ -1,12 +1,12 @@
|
|||
{
|
||||
"name": "@mariozechner/pi",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"lockfileVersion": 3,
|
||||
"requires": true,
|
||||
"packages": {
|
||||
"": {
|
||||
"name": "@mariozechner/pi",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"license": "MIT",
|
||||
"dependencies": {
|
||||
"@ai-sdk/openai": "^2.0.5",
|
||||
|
|
|
|||
|
|
@ -1,6 +1,6 @@
|
|||
{
|
||||
"name": "@mariozechner/pi",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"description": "CLI tool for managing vLLM deployments on GPU pods",
|
||||
"type": "module",
|
||||
"bin": {
|
||||
|
|
@ -34,7 +34,7 @@
|
|||
"node": ">=20.0.0"
|
||||
},
|
||||
"dependencies": {
|
||||
"@mariozechner/pi-agent": "^0.5.9",
|
||||
"@mariozechner/pi-agent": "^0.5.10",
|
||||
"chalk": "^5.5.0"
|
||||
},
|
||||
"devDependencies": {}
|
||||
|
|
|
|||
4
packages/tui/package-lock.json
generated
4
packages/tui/package-lock.json
generated
|
|
@ -1,12 +1,12 @@
|
|||
{
|
||||
"name": "@mariozechner/tui",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"lockfileVersion": 3,
|
||||
"requires": true,
|
||||
"packages": {
|
||||
"": {
|
||||
"name": "@mariozechner/tui",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"license": "MIT",
|
||||
"dependencies": {
|
||||
"@types/mime-types": "^2.1.4",
|
||||
|
|
|
|||
|
|
@ -1,6 +1,6 @@
|
|||
{
|
||||
"name": "@mariozechner/pi-tui",
|
||||
"version": "0.5.9",
|
||||
"version": "0.5.10",
|
||||
"description": "Terminal User Interface library with differential rendering for efficient text-based applications",
|
||||
"type": "module",
|
||||
"main": "dist/index.js",
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue