← Back to Models·

DeepSeek

DeepSeek
· 2026-05-01
DeepSeek's chat model. Outstanding reliability at the lowest cost, with moderate bearish tendency.
$Lowest Cost$0.0022/run
Highest Validity100.0%
DeepSeek is the most conservative model — smallest average valuation gap (-4.4%). Moderate daily volatility (2.20%/day). Wins: Lowest Cost, Highest Validity. Perfect validity rate (100%). Cheapest model at $0.0022/run.

Model Positioning

Calibration score vs estimate consistency
Calibrated & ConsistentConsistent but UncalibratedCalibrated but VolatileLow ScoresCalibrationConsistencyGPTCLAUDEGEMINIDEEPSEEKGROK
Compare:GPTClaudeGeminiGrok

Key Metrics

-4.4%
Avg Gap
57/100
Calibration
45%
Correction Rate
34%
Positive Bias %
Responsive
Volatility
44
Days Tracked

Bias History

Model bias over time vs other models

Prediction Accuracy

Directional accuracy vs actual price changes
Directional Accuracy
48.1%
Mean Absolute Error
19%
Prediction Bias
-4.5%
Hit Rate (±5%)
9.6%
0%25%50%75%100%3.3.11.3.19.3.27.3.6.4.14.4.22.4.30.4.
GPT51.5%
Claude49.6%
Gemini47.3%
DeepSeek48.1%
Grok49.3%
N = 989 · 1 Day

Current Estimates

Largest Positive Gap
+45.7%
+38.0%
+17.4%
+10.3%
+5.7%
Largest Negative Gap
-35.2%
-33.9%
-19.1%
-18.7%
-18.1%

Assumption Profile

6.4%
σ = 5.96
Revenue CAGR 5Y
9.6%
σ = 1.33
WACC
2.07%
σ = 0.20
Terminal Growth
23.5%
EBIT Margin Target

Validity

100.0%
1012 / 1012 validity rate

Cost & Performance

Avg Latency: 19.9s
Cost per Run: $0.0022
Total Cost: $2.23
~5495 + 658 tokens

Calibration Detail

57
/100
Correction Rate: 44.7%
Uncapped Deviation:
Total Valuations: 1012
AI Investor Barometer · DeepSeek← Back to Models
All content is generated by AI models and may contain errors. This is an experimental tool — not investment advice, research, or recommendation. Terms of Use · Privacy Policy