LLM Status
Model Availability
| Model | Status | Cost | Description | Used by |
|---|---|---|---|---|
| GEMINI | ||||
| gemini-2.5-flash |
Quota exhausted — resets in 13h24m56s
|
1x | Fast, good for simple tasks | l3_evaluate session_audit |
| gemini-2.5-pro |
Not probed
|
4x | Strong, good for most tasks | |
| gemini-3-flash-preview |
Quota exhausted — resets in 13h24m56s
|
1.3x | Gemini 3 Flash preview, fast | audit.meta audit.mcp_selfbuild audit.mcp_usage audit.knowledge audit.syncrovanis |
| gemini-3-pro-preview |
Not probed
|
5x | Gemini 3 Pro preview, best quality | |
| CODEX | ||||
| gpt-5.4 |
Not probed
|
6.3x | Latest frontier agentic coding model | |
| gpt-5.4-mini |
Available
|
1.9x | Smaller frontier agentic coding model | l3_classify expert_training |
| gpt-5.3-codex |
Not probed
|
5.6x | Frontier Codex-optimized agentic coding model | |
| gpt-5.2-codex |
Not probed
|
5.6x | Frontier agentic coding model | |
| gpt-5.2 |
Not probed
|
5.6x | Professional work and long-running agents | |
| gpt-5.1-codex-max |
Not probed
|
4x | Deep and fast reasoning, max context | |
| gpt-5.1-codex-mini |
Not probed
|
0.8x | Optimized for codex, cheaper and faster | |
| gpt-5.3-codex-spark |
Not probed
|
Text-only, near-instant coding iteration, Pro only | ||
| gpt-5.1-codex |
Not probed
|
4x | Code-optimized | |
| gpt-5.1 |
Not probed
|
4x | General purpose | |
| gpt-5-codex |
Not probed
|
4x | Older code-optimized | |
| gpt-5-codex-mini |
Not probed
|
0.8x | Older mini code-optimized | |
| gpt-5 |
Not probed
|
4x | General purpose | |
| gpt-5-mini |
Not probed
|
0.8x | Small, fast | |
| o3 |
Not probed
|
3.6x | Reasoning model, complex analysis | |
| o4-mini |
Not probed
|
2x | Fast reasoning model | |
| gpt-4.1 |
Not probed
|
3.6x | Legacy | |
| COPILOT | ||||
| gpt-5.4 |
Not probed
|
1x | GPT-5.4 default | |
| gpt-5.3-codex |
Not probed
|
1x | Codex via Copilot | |
| gpt-5.2-codex |
Not probed
|
1x | Older Codex via Copilot | |
| gpt-5.2 |
Not probed
|
1x | GPT via Copilot | |
| gpt-5.1-codex-max |
Not probed
|
1x | Max context via Copilot | |
| gpt-5.1-codex |
Not probed
|
1x | Codex via Copilot | |
| gpt-5.1 |
Not probed
|
1x | GPT via Copilot | |
| gpt-5.4-mini |
Not probed
|
0.33x | Mini GPT via Copilot | |
| gpt-5.1-codex-mini |
Not probed
|
0.33x | Mini Codex preview via Copilot | |
| gpt-5-mini |
Not probed
|
0x | Mini GPT via Copilot | |
| gpt-4.1 |
Not probed
|
0x | Legacy GPT via Copilot | |
| claude-sonnet-4.6 |
Not probed
|
1x | Strong Claude | |
| claude-sonnet-4.5 |
Not probed
|
1x | Previous Sonnet | |
| claude-haiku-4.5 |
Not probed
|
0.33x | Fast/cheap Claude | |
| claude-opus-4.6 |
Not probed
|
3x | Best quality | |
| claude-opus-4.5 |
Not probed
|
3x | Previous Opus | |
| claude-sonnet-4 |
Not probed
|
1x | Older Sonnet | |
| gemini-3-pro-preview |
Not probed
|
1x | Gemini 3 Pro via Copilot | |
| CLAUDE | ||||
| haiku |
Not probed
|
1x | Fast, cheap inline subagent | |
| sonnet |
Not probed
|
5x | Balanced inline subagent | avslutt_implement |
| opus |
Not probed
|
25x | Most capable inline subagent | |
Pipeline Configuration
Edit in Config
| Pipeline Step | Provider : Model | Effort | Health |
|---|---|---|---|
| l3_classify | codex:gpt-5.4-mini | ||
| l3_evaluate | gemini:gemini-2.5-flash | high | |
| expert_training | codex:gpt-5.4-mini | medium | |
| session_audit | gemini:gemini-2.5-flash | ||
| meta | gemini:gemini-3-flash-preview | ||
| mcp_selfbuild | gemini:gemini-3-flash-preview | ||
| mcp_usage | gemini:gemini-3-flash-preview | ||
| knowledge | gemini:gemini-3-flash-preview | ||
| syncrovanis | gemini:gemini-3-flash-preview | ||
| formatter | inherits parent | — | |
| avslutt_implement | claude:sonnet | high |