Foresight-v3
Better AI Predictions.
Better accuracy.
Cheaper inference.
OpenAI-compatible API.
ProphetArena 3/25/2026
Trusted for high-stakes predictions
Verified edge.
Foresight-v3 is #1 in ProphetArena Sports and Politics, ahead of frontier models, and the only model with a positive edge in both categories.
| Rank | Org | Model | Type | Events | Brier Score | Score vs Market | Avg Return | ECE |
|---|---|---|---|---|---|---|---|---|
| 1 |  | Foresight V3 | Fixed Context | 1,000 | 0.9315 | +0.0729 | 83.51% | 0.0815 |
| 2 |  | Gemini 3 Pro | Fixed Context | 2,324 | 0.9105 | +0.0368 | 85.92% | 0.0369 |
| 3 |  | GPT-5.2 | Fixed Context | 4,228 | 0.9094 | +0.0286 | 90.14% | 0.0329 |
| 4 | | GPT-5.2 high | Fixed Context | 4,354 | 0.9085 | +0.0276 | 88.01% | 0.0293 |
| 5 |  | Claude Opus 4.5 | Fixed Context | 4,216 | 0.9079 | +0.0267 | 81.79% | 0.0350 |
ProphetArena overall, 3/25/2026
| Rank | Org | Model | Type | Events | 1 - Brier Score | Edge Over Market | Avg Return | ECE |
|---|---|---|---|---|---|---|---|---|
| 1 | | Foresight V3 | Fixed Context | 1,000 | 0.9588 (±0.0067) | +0.0586 | 106.67% (±17.07) | 0.0504 |
| 2 |  | Grok 4.1 Fast | Fixed Context | 3,895 | 0.9249 (±0.0043) | +0.0391 | 94.58% (±8.19) | 0.0312 |
| 3 | | GPT-5.2 high | Fixed Context | 4,354 | 0.9269 (±0.0041) | +0.0388 | 91.74% (±8.40) | 0.0233 |
| 4 | | GPT-5.2 | Fixed Context | 4,228 | 0.9263 (±0.0041) | +0.0382 | 94.63% (±9.77) | 0.0279 |
| 5 |  | DeepSeek V3.2 | Fixed Context | 4,307 | 0.9233 (±0.0045) | +0.0372 | 86.15% (±7.82) | 0.0367 |
ProphetArena sports, 5/13/2026
| Rank | Org | Model | Type | Events | 1 - Brier Score | Edge Over Market | Avg Return | ECE |
|---|---|---|---|---|---|---|---|---|
| 1 | | Foresight V3 | Fixed Context | 1,000 | 0.9805 (±0.0189) | +0.0804 | 136.50% (±35.24) | 0.0504 |
| 2 | | Gemini 3 Pro | Fixed Context | 2,324 | 0.9334 (±0.0205) | +0.0591 | 91.12% (±25.71) | 0.0344 |
| 3 | | GPT-5.2 high | Fixed Context | 4,354 | 0.9312 (±0.0203) | +0.0432 | 92.01% (±24.46) | 0.0233 |
| 4 | | GPT-5.2 | Fixed Context | 4,228 | 0.9206 (±0.0211) | +0.0325 | 75.58% (±23.78) | 0.0279 |
| 5 | | Grok 4.1 Fast | Fixed Context | 3,895 | 0.9177 (±0.0269) | +0.0319 | 82.91% (±28.29) | 0.0312 |
ProphetArena politics, 5/13/2026
OpenAI Compatible API
Drop calibrated forecasts into agents, notebooks, market tooling, or research pipelines.
client = OpenAI( api_key="lightningrod-api-key", base_url="https://api.lightningrod.ai/api/public/v1/openai", ) question = "Will the Fed cut rates in March 2026?" prediction = client.chat.completions.create( model="LightningRodLabs/foresight-v3", messages=[{"role": "user", "content": question}], )