docs(guides): add common provider combinations section (#13735) (#15195)

Validated in a combined board with the other PRs of this wave on release/v3.8.52: the PR's own tests pass; typecheck:core, open-sse typecheck, api-typecheck, complexity, changelog-integrity, env-doc-sync, test-discovery, vitest-exclusions, error-helper, test-masking, lockfile, deps, gate-manifest, cli-i18n and i18n coverage green. Any file-size ceiling or stryker registration this wave needs lands in the wave's follow-up PR. Thank you for the contribution!
This commit is contained in:
Ronaldo Davi 2026-10-08 02:01:51 -03:00 • committed by GitHub
parent d7070a5e96
commit c69dce259f
No known key found for this signature in database
GPG key ID: B5690EEEBB952194

View file

@ -911,6 +911,65 @@ Chain: production-fallback
---
### Common Provider Combinations & Routing Patterns
Here are example patterns for combining multiple providers and routing between them in OmniRoute:
#### 1. Coding Agent Combo: High-End Reasoning with Cost/Speed Fallback
Ideal for coding agents (OpenCode, Claude Code, Cursor, Cline). Routes initially to frontier reasoning models, falling back to fast coding models on quota exhaustion or errors.
- **Dashboard**: Combos → New Combo → Name: `agent-coding` → Strategy: `Priority`
- **Models**:
1. `claude/claude-sonnet-4-6` (Primary coding agent)
2. `openai/gpt-4o` (Secondary high-capacity fallback)
3. `deepseek/deepseek-v4-flash` (High-efficiency, cost-effective fallback)
```bash
# Example via API
curl -X POST http://localhost:20128/api/combos \
-H "Authorization: Bearer $OMNIROUTE_MANAGE_KEY" \
-H "Content-Type: application/json" \
-d '{
"name": "agent-coding",
"strategy": "priority",
"models": [
{ "model": "claude/claude-sonnet-4-6" },
{ "model": "openai/gpt-4o" },
{ "model": "deepseek/deepseek-v4-flash" }
]
}'
```
#### 2. Free-Tier Auto Failover Combo
Chains multiple free-tier and keyless providers to maximize uptime without API costs.
- **Strategy**: `Least Used` or `Round Robin` (distributes load across quotas)
- **Models**:
1. `groq/llama-3.3-70b-versatile`
2. `gemini/gemini-2.5-flash`
3. `cerebras/gpt-oss-120b`
```bash
# Example via CLI
omniroute combo create free-stack \
--models "groq/llama-3.3-70b-versatile,gemini/gemini-2.5-flash,cerebras/gpt-oss-120b" \
--strategy round-robin
```
#### 3. Multimodal / Vision & Text Pipeline
Pairing specialized vision models with high-speed text generation for workflows involving image understanding and code generation.
- **Pattern**: A `Priority` combo that lists vision-capable models first and a high-throughput text/code model last.
- **Models**:
1. `gemini/gemini-2.5-pro` (Strong image/multimodal comprehension)
2. `openai/gpt-4o` (Balanced vision and tool usage)
3. `deepseek/deepseek-v4-flash` (Text/code generation)
---
### Resilience & Circuit Breakers
Configure via **Dashboard → Settings → Resilience**.