Skip to content

Garry's Opinionated OpenClaw/Hermes Agent Brain

TypeScript
2.9 KB
Raw Blame
import type { Recipe } from '../types.ts';

/**
 * Alibaba DashScope (灵积). OpenAI-compatible /embeddings endpoint at
 * dashscope-intl.aliyuncs.com. Hosts text-embedding-v2 (older) and
 * text-embedding-v3 (current; Matryoshka-aware up to 1024 dims).
 *
 * Reference: https://help.aliyun.com/zh/model-studio/getting-started/
 *
 * Note: the international endpoint requires a region-aware DASHSCOPE_API_KEY.
 * China-region users typically point at https://dashscope.aliyuncs.com/...
 * via cfg.base_urls['dashscope']. v0.32 ships with the international
 * default; users override per the recipe convention.
 */
export const dashscope: Recipe = {
  id: 'dashscope',
  name: 'Alibaba DashScope (灵积)',
  tier: 'openai-compat',
  implementation: 'openai-compatible',
  // No shipped rate for these yet: they warn and run under a default cap; a
  // user cap refuses them until `gbrain pricing set` registers one (X1 guard).
  unpriced_models: [
    'text-embedding-v3',
    'text-embedding-v2',
    'text-embedding-v1',
  ],
  base_url_default: 'https://dashscope-intl.aliyuncs.com/compatible-mode/v1',
  auth_env: {
    required: ['DASHSCOPE_API_KEY'],
    setup_url: 'https://help.aliyun.com/zh/model-studio/getting-started/',
  },
  touchpoints: {
    embedding: {
      models: ['text-embedding-v3', 'text-embedding-v2', 'text-embedding-v1'],
      default_dims: 1024,
      // PR #5922 / #4530: DashScope's documented per-INPUT caps (Model Studio
      // "Synchronous API for text embedding", read 2026-10-11): v1/v2 reject
      // any single input over 2,048 tokens with a 400 that fails the whole
      // batch; v3 accepts 8,192. The chunker splits to floor(cap × 0.6)
      // (resolveMaxChunkTokens) and the heal pass splits stored rows.
      max_input_tokens: {
        'text-embedding-v1': 2048,
        'text-embedding-v2': 2048,
        'text-embedding-v3': 8192,
      },
      dims_options: [64, 128, 256, 512, 768, 1024],
      // Alibaba doesn't publish a hard batch-token cap for the OpenAI-compat
      // path. Conservative declaration so the gateway pre-splits before
      // hitting whatever undocumented server-side limit exists.
      max_batch_tokens: 8192,
      // DashScope's OpenAI-compat /embeddings endpoint rejects requests with
      // more than 10 input items (documented Model Studio cap). The gateway's
      // capBatchItems pre-split enforces this; max_batch_tokens above keeps
      // guarding aggregate token size. Concept from community PRs #2643/#2405.
      max_batch_items: 10,
      // text-embedding-v3 mixes English + CJK heavily; the tokenizer is
      // closer to Voyage density than OpenAI tiktoken for CJK-dominant
      // content. Conservative chars_per_token=2 leaves headroom.
      chars_per_token: 2,
    },
  },
  setup_hint:
    'Get an API key at https://help.aliyun.com/zh/model-studio/getting-started/, then `export DASHSCOPE_API_KEY=...`',
};