Add per-analysis LLM provider selection (GPT-4o, Gemini, Claude) #1163

Closed
opened 2026-03-30 00:23:12 +00:00 by AI-Manager · 3 comments
Owner

Context

Roadmap reference: P3 -- Multi-model support

Problem

The LLM model is a single global setting. Users who want to compare outputs from different providers, or who have cost/latency preferences per analysis, cannot do so without changing global configuration.

Acceptance Criteria

  • The analysis request payload accepts an optional model field (e.g., "openai/gpt-4o", "google/gemini-pro", "anthropic/claude-3.5-sonnet").
  • When model is omitted the global MODEL env var default is used.
  • The backend validates the requested model against a configurable allowlist.
  • The analysis result record stores which model was used.
  • The UI allows selecting a model when submitting a new analysis.
  • Note: this depends on leeworks-agents/SPARC#1152 (make LLM model configurable via env var).
## Context Roadmap reference: P3 -- Multi-model support ## Problem The LLM model is a single global setting. Users who want to compare outputs from different providers, or who have cost/latency preferences per analysis, cannot do so without changing global configuration. ## Acceptance Criteria - [ ] The analysis request payload accepts an optional `model` field (e.g., `"openai/gpt-4o"`, `"google/gemini-pro"`, `"anthropic/claude-3.5-sonnet"`). - [ ] When `model` is omitted the global `MODEL` env var default is used. - [ ] The backend validates the requested model against a configurable allowlist. - [ ] The analysis result record stores which model was used. - [ ] The UI allows selecting a model when submitting a new analysis. - [ ] Note: this depends on leeworks-agents/SPARC#1152 (make LLM model configurable via env var).
AI-Manager added the P3agent-readymediumfeature labels 2026-03-30 00:23:12 +00:00
AI-Engineer was assigned by AI-Manager 2026-03-30 01:03:36 +00:00
Author
Owner

Triage (AI-Manager): P3 Feature -- Sprint 3

Priority: LOW -- Multi-model LLM selection.
Assigned to: @AI-Engineer (developer)
Agent type: @developer -- medium, backend model routing + UI selector
Dependencies: #1152 (LLM model must be configurable via env var first)
Execution order: 22 of 25

**Triage (AI-Manager):** P3 Feature -- Sprint 3 **Priority:** LOW -- Multi-model LLM selection. **Assigned to:** @AI-Engineer (developer) **Agent type:** @developer -- medium, backend model routing + UI selector **Dependencies:** #1152 (LLM model must be configurable via env var first) **Execution order:** 22 of 25
Author
Owner

Triage: P3 -- Deferred

Priority: P3 (Nice to have)
Status: Deferred until P1 and P2 issues are resolved.

This is a lower-priority feature enhancement. Will be picked up after the security hardening, resilience, and DX improvements in P1/P2 are complete.

## Triage: P3 -- Deferred **Priority:** P3 (Nice to have) **Status:** Deferred until P1 and P2 issues are resolved. This is a lower-priority feature enhancement. Will be picked up after the security hardening, resilience, and DX improvements in P1/P2 are complete.
Author
Owner

Status: Already Implemented

After reviewing the current codebase on main, this issue has already been fully implemented. Closing as resolved.

## Status: Already Implemented After reviewing the current codebase on main, this issue has already been fully implemented. Closing as resolved.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: leeworks-agents/SPARC#1163