Skip to content

Commit

Permalink
chore: models update 2024-08-18 (#59)
Browse files Browse the repository at this point in the history
  • Loading branch information
@mdjastrzebski
mdjastrzebski authored Aug 18, 2024
1 parent 98d9986 commit 1b11822
Show file tree
Hide file tree
Showing 11 changed files with 2,392 additions and 2,291 deletions.
894 changes: 0 additions & 894 deletions .yarn/releases/yarn-4.2.1.cjs
View file

This file was deleted.

925 changes: 925 additions & 0 deletions .yarn/releases/yarn-4.4.0.cjs
View file

Large diffs are not rendered by default.

2 changes: 1 addition & 1 deletion .yarnrc.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,4 +6,4 @@ nmHoistingLimits: workspaces

nodeLinker: node-modules

yarnPath: .yarn/releases/yarn-4.2.1.cjs
yarnPath: .yarn/releases/yarn-4.4.0.cjs
10 changes: 5 additions & 5 deletions package.json
View file
Original file line number Diff line number Diff line change
Expand Up @@ -41,20 +41,20 @@
"registry": "https://registry.npmjs.org/"
},
"dependencies": {
"@anthropic-ai/sdk": "^0.24.3",
"@anthropic-ai/sdk": "^0.26.1",
"@inkjs/ui": "^1.0.0",
"@mistralai/mistralai": "^0.5.0",
"@mistralai/mistralai": "^1.0.2",
"chalk": "^5.3.0",
"date-fns": "^3.6.0",
"dotenv": "^16.4.5",
"ink": "^4.4.1",
"ink-link": "^3.0.0",
"ink-text-input": "^5.0.1",
"openai": "^4.52.2",
"openai": "^4.56.0",
"prompts": "^2.4.2",
"react": "^18.2.0",
"redent": "^4.0.0",
"tiktoken": "^1.0.15",
"tiktoken": "^1.0.16",
"update-notifier": "^7.0.0",
"yargs": "^17.7.2",
"zod": "^3.23.6",
Expand All @@ -80,7 +80,7 @@
"release-it": "^15.11.0",
"typescript": "^5.4.5"
},
"packageManager": "yarn@4.2.1",
"packageManager": "yarn@4.4.0",
"engines": {
"node": ">= 18.0.0"
},
Expand Down
2 changes: 1 addition & 1 deletion src/commands/chat/state/init.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -23,7 +23,7 @@ export function initChatState(

const modelOrAlias = options.model ?? providerFileConfig.model;
const model = modelOrAlias
? provider.modelAliases[modelOrAlias] ?? modelOrAlias
? (provider.modelAliases[modelOrAlias] ?? modelOrAlias)
: provider.defaultModel;

const systemPrompt = providerFileConfig.systemPrompt ?? DEFAULT_SYSTEM_PROMPT;
Expand Down
5 changes: 4 additions & 1 deletion src/engine/providers/anthropic.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -21,10 +21,13 @@ const Anthropic: Provider = {
// Price per 1M tokens [input, output].
// Source: https://docs.anthropic.com/en/docs/about-claude/models
modelPricing: {
'claude-3-haiku-20240307': { inputTokensCost: 0.25, outputTokensCost: 1.25 },
// Current models
'claude-3-5-sonnet-20240620': { inputTokensCost: 3.0, outputTokensCost: 15.0 },
'claude-3-haiku-20240307': { inputTokensCost: 0.25, outputTokensCost: 1.25 },
'claude-3-sonnet-20240229': { inputTokensCost: 3.0, outputTokensCost: 15.0 },
'claude-3-opus-20240229': { inputTokensCost: 15.0, outputTokensCost: 75.0 },

// Legacy models
'claude-2.1': { inputTokensCost: 8.0, outputTokensCost: 24.0 },
'claude-2.0': { inputTokensCost: 8.0, outputTokensCost: 24.0 },
'claude-instant-1.2': { inputTokensCost: 0.8, outputTokensCost: 2.4 },
Expand Down
49 changes: 26 additions & 23 deletions src/engine/providers/mistral.ts
View file
Original file line number Diff line number Diff line change
@@ -1,4 +1,5 @@
import MistralClient, { type ChatCompletionResponseChunk } from '@mistralai/mistralai';
import { Mistral as MistralClient } from '@mistralai/mistralai';
import type { CompletionEvent } from '@mistralai/mistralai/models/components/completionevent.js';
import { type Message, type ModelResponseUpdate } from '../inference.js';
import { estimateInputTokens, estimateOutputTokens } from '../tokenizer.js';
import { responseStyles, type ProviderConfig } from './config.js';
Expand All @@ -9,54 +10,56 @@ const Mistral: Provider = {
name: 'mistral',
apiKeyUrl: 'https://console.mistral.ai/api-keys/',

// OpenAI models: https://docs.mistral.ai/platform/endpoints/
// Mistral models: https://docs.mistral.ai/getting-started/models/
defaultModel: 'mistral-large-latest',

// Price per 1M tokens [input, output].
// Source: https://docs.mistral.ai/platform/pricing/
modelPricing: {
// Current models
'open-mistral-nemo': { inputTokensCost: 0.3, outputTokensCost: 0.3 },
'open-mistral-nemo-2407': { inputTokensCost: 0.3, outputTokensCost: 0.3 },
'mistral-large-latest': { inputTokensCost: 3, outputTokensCost: 9 },
'mistral-large-2407': { inputTokensCost: 3, outputTokensCost: 9 },
'codestral-latest': { inputTokensCost: 1, outputTokensCost: 3 },
'codestral-2405': { inputTokensCost: 1, outputTokensCost: 3 },

// Legacy models
'open-mistral-7b': { inputTokensCost: 0.25, outputTokensCost: 0.25 },
'open-mixtral-8x7b': { inputTokensCost: 0.7, outputTokensCost: 0.7 },
'open-mixtral-8x22b': { inputTokensCost: 2, outputTokensCost: 6 },
'mistral-small-latest': { inputTokensCost: 1, outputTokensCost: 3 },
'mistral-small-2402': { inputTokensCost: 1, outputTokensCost: 3 },
'mistral-medium-latest': { inputTokensCost: 2.7, outputTokensCost: 8.1 },
'mistral-medium-2312': { inputTokensCost: 2.7, outputTokensCost: 8.1 },
'mistral-large-latest': { inputTokensCost: 4, outputTokensCost: 12 },
'mistral-large-2402': { inputTokensCost: 4, outputTokensCost: 12 },
'codestral-latest': { inputTokensCost: 1, outputTokensCost: 3 },
'codestral-2405': { inputTokensCost: 1, outputTokensCost: 3 },
},

modelAliases: {
mistral: 'open-mistral-7b',
mixtral: 'open-mixtral-8x22b',
small: 'mistral-small-latest',
medium: 'mistral-medium-latest',
nemo: 'open-mistral-nemo-2407',
large: 'mistral-large-latest',
codestral: 'codestral-latest',
},

getChatCompletion: async (config: ProviderConfig, messages: Message[]) => {
const api = new MistralClient(config.apiKey);
const api = new MistralClient({ apiKey: config.apiKey });
const allMessages = getMessages(config, messages);

const startTime = performance.now();
const response = await api.chat({
messages: allMessages,
const response = await api.chat.complete({
model: config.model,
messages: allMessages,
...getMistralResponseStyle(config),
});
const responseTime = performance.now() - startTime;

return {
message: {
role: 'assistant',
content: response.choices[0]?.message.content ?? '',
content: response.choices?.[0]?.message?.content ?? '',
},
usage: {
inputTokens: response.usage?.prompt_tokens ?? 0,
outputTokens: response.usage?.completion_tokens ?? 0,
inputTokens: response.usage.promptTokens,
outputTokens: response.usage.completionTokens,
requests: 1,
},
responseTime,
Expand All @@ -70,21 +73,21 @@ const Mistral: Provider = {
messages: Message[],
onResponseUpdate: (update: ModelResponseUpdate) => void,
) {
const api = new MistralClient(config.apiKey);
const api = new MistralClient({ apiKey: config.apiKey });
const allMessages = getMessages(config, messages);

const startTime = performance.now();
const stream = await api.chatStream({
const stream = await api.chat.stream({
messages: allMessages,
model: config.model,
...getMistralResponseStyle(config),
});

let lastChunk: ChatCompletionResponseChunk | null = null;
let lastChunk: CompletionEvent | null = null;
let content = '';
for await (const chunk of stream) {
lastChunk = chunk;
content += chunk.choices[0]?.delta?.content || '';
content += chunk.data.choices[0]?.delta?.content || '';
onResponseUpdate({ content });
}

Expand All @@ -96,12 +99,12 @@ const Mistral: Provider = {
content,
},
usage: {
inputTokens: lastChunk?.usage?.prompt_tokens ?? estimateInputTokens(allMessages),
outputTokens: lastChunk?.usage?.completion_tokens ?? estimateOutputTokens(content),
inputTokens: lastChunk?.data.usage?.promptTokens ?? estimateInputTokens(allMessages),
outputTokens: lastChunk?.data.usage?.completionTokens ?? estimateOutputTokens(content),
requests: 1,
},
responseTime,
responseModel: lastChunk?.model || 'unknown',
responseModel: lastChunk?.data.model || 'unknown',
data: lastChunk,
};
},
Expand Down
11 changes: 6 additions & 5 deletions src/engine/providers/open-ai.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -10,12 +10,16 @@ const OpenAi: Provider = {
apiKeyUrl: 'https://platform.openai.com/api-keys',

// OpenAI models: https://platform.openai.com/docs/models
defaultModel: 'gpt-4o',
defaultModel: 'gpt-4o-2024-08-06',

// Price per 1M tokens [input, output].
// Source: https://openai.com/pricing
modelPricing: {
'gpt-4o': { inputTokensCost: 5, outputTokensCost: 15 },
'gpt-4o-2024-08-06': { inputTokensCost: 2.5, outputTokensCost: 10 },
'gpt-4o-2024-05-13': { inputTokensCost: 5, outputTokensCost: 15 },
'gpt-4o-mini': { inputTokensCost: 0.15, outputTokensCost: 0.6 },
'gpt-4o-mini-2024-07-18': { inputTokensCost: 0.15, outputTokensCost: 0.6 },
'gpt-4-turbo': { inputTokensCost: 10, outputTokensCost: 30 },
'gpt-4-turbo-2024-04-09': { inputTokensCost: 10, outputTokensCost: 30 },
'gpt-4': { inputTokensCost: 30, outputTokensCost: 60 },
Expand All @@ -25,10 +29,7 @@ const OpenAi: Provider = {
'gpt-3.5-turbo-instruct': { inputTokensCost: 1.5, outputTokensCost: 2.0 },
},

modelAliases: {
'gpt-4-turbo-preview': 'gpt-4-turbo',
'gpt-3.5': 'gpt-3.5-turbo',
},
modelAliases: {},

getChatCompletion: async (config: ProviderConfig, messages: Message[]) => {
const api = new OpenAI({
Expand Down
28 changes: 15 additions & 13 deletions src/engine/providers/perplexity.ts
View file
Original file line number Diff line number Diff line change
Expand Up @@ -10,35 +10,37 @@ const Perplexity: Provider = {
apiKeyUrl: 'https://perplexity.ai/settings/api',

// Perplexity models: https://docs.perplexity.ai/docs/model-cards
defaultModel: 'llama-3-sonar-large-32k-chat',
defaultModel: 'llama-3.1-sonar-huge-128k-online',

// Price per 1M tokens [input, output], per 1k requests.
// Source: https://docs.perplexity.ai/docs/model-cards
// Source: https://docs.perplexity.ai/docs/pricing
modelPricing: {
'llama-3-sonar-small-32k-chat': { inputTokensCost: 0.2, outputTokensCost: 0.2 },
'llama-3-sonar-small-32k-online': {
'llama-3.1-sonar-small-128k-online': {
inputTokensCost: 0.2,
outputTokensCost: 0.2,
requestsCost: 5,
},
'llama-3-sonar-large-32k-chat': { inputTokensCost: 1, outputTokensCost: 1 },
'llama-3-sonar-large-32k-online': {
'llama-3.1-sonar-large-128k-online': {
inputTokensCost: 1,
outputTokensCost: 1,
requestsCost: 5,
},
'llama-3-8b-instruct': { inputTokensCost: 0.2, outputTokensCost: 0.2 },
'llama-3-70b-instruct': { inputTokensCost: 1, outputTokensCost: 1 },
'mixtral-8x7b-instruct': { inputTokensCost: 0.6, outputTokensCost: 0.6 },
'llama-3.1-sonar-huge-128k-online': {
inputTokensCost: 5,
outputTokensCost: 5,
requestsCost: 5,
},
'llama-3.1-sonar-small-128k-chat': { inputTokensCost: 0.2, outputTokensCost: 0.2 },
'llama-3.1-sonar-large-128k-chat': { inputTokensCost: 1, outputTokensCost: 1 },
'llama-3.1-8b-instruct': { inputTokensCost: 0.2, outputTokensCost: 0.2 },
'llama-3.1-70b-instruct': { inputTokensCost: 1, outputTokensCost: 1 },
},

modelAliases: {
'small': 'llama-3-sonar-small-32k-chat',
'large': 'llama-3-sonar-large-32k-chat',
'online': 'llama-3-sonar-large-32k-online',
'llama-3': 'llama-3-70b-instruct',
'mixtral': 'mixtral-8x7b-instruct',
small: 'llama-3.1-sonar-small-128k-online',
large: 'llama-3.1-sonar-large-128k-online',
huge: 'llama-3.1-sonar-huge-128k-online',
},

getChatCompletion: async (config: ProviderConfig, messages: Message[]) => {
Expand Down
56 changes: 26 additions & 30 deletions website/docs/getting-started.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -68,30 +68,29 @@ CLI options are passed when invoking the `ai` commend:

You should have a relevant API key in your `~/.airc.json` file.

### Models
### Current Models

<Tabs groupId="provider">
<TabItem value="openAi" label="Open AI">

| Model | Alias | Price: in \| out \* | Notes |
| --------------- | --------- | ------------------- | ------- |
| `gpt-4o` | | $5 \| $15 | Default |
| `gpt-4-turbo` | | $10 \| $30 | |
| `gpt-4` | | $30 \| $60 | |
| `gpt-3.5-turbo` | `gpt-3.5` | $0.5 \| $1.5 | |
| Model | Alias | Price: in \| out \* | Notes |
| ------------------- | ----- | ------------------- | ------- |
| `gpt-4o-2024-08-06` | | $2.5 \| $10 | Default |
| `gpt-4o` | | $5 \| $15 | |
| `gpt-4o-mini` | | $0.15 \| $0.6 | |

\* API prices per 1 million input/output tokens

More info: [OpenAI docs](https://platform.openai.com/docs/models)
More models & info: [OpenAI docs](https://platform.openai.com/docs/models)

</TabItem>
<TabItem value="anthropic" label="Anthropic">

| Model | Alias | Price: in \| out \* | Notes |
| -------------------------- | -------- | ------------------- | ------- |
| `claude-3-opus-20240229` | `opus` | $15 \| $75 | |
| `claude-3-sonnet-20240229` | `sonnet` | $3 \| $15 | Default |
| `claude-3-haiku-20240307` | `haiku` | $0.25 \| $1.25 | |
| Model | Alias | Price: in \| out \* | Notes |
| ---------------------------- | -------- | ------------------- | ------- |
| `claude-3-5-sonnet-20240620` | `sonnet` | $3 \| $15 | Default |
| `claude-3-opus-20240229` | `opus` | $15 \| $75 | |
| `claude-3-haiku-20240307` | `haiku` | $0.25 \| $1.25 | |

\* API prices per 1 million input/output tokens

Expand All @@ -100,15 +99,15 @@ More info: [Anthropic docs](https://docs.anthropic.com/claude/docs/models-overvi
</TabItem>
<TabItem value="perplexity" label="Perplexity">

| Model | Alias | Price: in \| out \* | Notes |
| -------------------------------- | --------- | --------------------------- | ---------- |
| `llama-3-sonar-large-32k-chat` | `large` | $1 \| $1 | Default |
| `llama-3-sonar-large-32k-online` | `online` | $1 \| $1 \| request: $5 | Online\*\* |
| `llama-3-sonar-small-32k-chat` | `small` | $0.2 \| $0.2 | |
| `llama-3-sonar-small-32k-online` | | $0.2 \| $0.2 \| request: $5 | Online\*\* |
| `llama-3-70b-instruct` | `llama-3` | $1 \| $1 | |
| `llama-3-8b-instruct` | | $0.2 \| $0.2 | |
| `mixtral-8x7b-instruct` | `mixtral` | $0.6 \| $0.6 | |
| Model | Alias | Price: in \| out \* | Notes |
| ----------------------------------- | ------- | ---------------------------- | ------------------- |
| `llama-3.1-sonar-huge-128k-online` | `huge` | $5 \| $5 \| requests: $5 | Default, Online\*\* |
| `llama-3.1-sonar-large-128k-online` | `large` | $1 \| $1 \| requests: $5 | Online\*\* |
| `llama-3.1-sonar-small-128k-online` | `small` | $0.2 \| $0.2 \| requests: $5 | Online\*\* |
| `llama-3.1-sonar-large-128k-chat` | | $1 \| $1 | |
| `llama-3.1-sonar-small-128k-chat` | | $0.2 \| $0.2 | |
| `llama-3.1-70b-instruct` | | $1 \| $1 | |
| `llama-3.1-8b-instruct` | | $0.2 \| $0.2 | |

\* API prices per 1 million input/output tokens, per 1 thousands requests

Expand All @@ -119,14 +118,11 @@ More info: [Perplexity docs](https://docs.perplexity.ai/docs/model-cards)
</TabItem>
<TabItem value="mistral" label="Mistral">

| Model | Alias | Price: in \| out \* | Notes |
| ----------------------- | --------- | ------------------- | ------- |
| `mistral-large-latest` | `large` | $4 \| $12 | Default |
| `mistral-medium-latest` | `medium` | $2.7 \| $8.1 | |
| `mistral-small-latest` | `small` | $1 \| $3 | |
| `open-mixtral-8x22b` | `mixtral` | $2 \| $6 | |
| `open-mixtral-8x7b` | | $0.7 \| $0.7 | |
| `open-mistral-7b` | `mistral` | $0.25 \| $0.25 | |
| Model | Alias | Price: in \| out \* | Notes |
| ---------------------- | ----------- | ------------------- | ------- |
| `mistral-large-latest` | `large` | $3 \| $9 | Default |
| `open-mistral-nemo` | `nemo` | $0.3 \| $0.3 | |
| `codestral-latest` | `codestral` | $1 \| $3 | |

\* API prices per 1 million input/output tokens

Expand Down
Loading

0 comments on commit 1b11822

Please sign in to comment.