← Back to dashboard·

DeepSeek

DeepSeek
· 2026-08-28
DeepSeek's chat model. Outstanding reliability at the lowest cost, with moderate bearish tendency.
Calibration and accuracy describe observed behavior, not predictive ability. Experimental research — not investment advice. Terms of Use
$Lowest Cost$0.0034/run
~Market Pulse-1.2%
DeepSeek is a moderate model with an average valuation gap of -1.2%. Moderate daily volatility (2.70%/day). Wins: Lowest Cost, Market Pulse. High validity rate (99.0%). Cheapest model at $0.0034/run.

Model Positioning

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

Key Metrics

-1.2%
Avg Gap
48/100
Calibration
14%
Correction Rate
45%
Positive Bias %
Responsive
Volatility
33
Days Tracked

Bias History

Model bias over time vs other models

Prediction Accuracy

Directional accuracy vs actual price changes
Directional Accuracy
52.1%
Mean Absolute Error
30.8%
Prediction Bias
-1.4%
Hit Rate (±5%)
12.5%
v8.10%25%50%75%100%14.7.21.7.28.7.4.8.12.8.19.8.27.8.DeepSeekClaudeGeminiGrokGPT
N = 746 · 1 Day

Current Estimates

Largest Positive Gap
+94.7%
+83.4%
+52.2%
+38.4%
+26.6%
Largest Negative Gap
-62.3%
-56.2%
-45.7%
-38.4%
-33.9%

Assumption Profile

6.6%
σ = 5.60
Revenue CAGR 5Y
9.4%
σ = 1.34
WACC
2.04%
σ = 0.15
Terminal Growth
23.4%
EBIT Margin Target

Validity

99.0%
2955 / 2984 validity rate

Cost & Performance

Avg Latency: 19.4s
Cost per Run: $0.0034
Total Cost: $10.00
~5763 + 1650 tokens

Calibration Detail

48
/100
Correction Rate: 13.5%
Uncapped Deviation:
Total Valuations: 770
Want these insights weekly?
Subscribe to AI Signals →
AI Investor Barometer · DeepSeek← Back to dashboard
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