ModelVerdict

Comparison

GPT-6 Sol vs Mistral Large 4

Same documents, same prompt: where each model gets it right, what it really costs, and how fast it responds.

Verdict

GPT-6 Sol is more accurate and cheaper in 23 of 30 combinations.

More accurate

GPT-6 Sol

wins 26 of 30

Cheaper

GPT-6 Sol

cheaper in 23 of 30

Faster

GPT-6 Sol

2.9 sec vs 31.5 sec

Fewer fixes

GPT-6 Sol

13 vs 107 documents per 1,000

Across 30 task and document-language combinations (speed and fixes: median).

At a glance

English (US)
Accuracy (correct fields)Per 1,000 documents
Task · documentsGPT-6 SolMistral Large 4DifferenceGPT-6 SolMistral Large 4Difference
Invoice extraction99.94%95.39%▼ 4.55$9.65$6.65−31 %
Email triage99.04%98.74%▼ 0.3$2.44$3.52+44 %
Personal data detection99.71%99.62%▼ 0.09$2.09$2.89+38 %
Contract clauses100.00%96.10%▼ 3.9$3.33$5.24+57 %
Document summaries99.56%100.00%▲ 0.44$3.85$5.62+46 %
Questions about a document100.00%98.89%▼ 1.11$2.36$4.16+76 %
Czech
Accuracy (correct fields)Per 1,000 documents
Task · documentsGPT-6 SolMistral Large 4DifferenceGPT-6 SolMistral Large 4Difference
Invoice extraction99.90%97.08%▼ 2.82$10.12$5.69−44 %
Email triage98.96%98.15%▼ 0.81$3.19$4.41+38 %
Personal data detection99.71%99.33%▼ 0.38$2.22$3.15+42 %
Contract clauses100.00%100.00%±0$4.62$4.58−1 %
Document summaries100.00%83.33%▼ 16.67$4.72$10.72+127 %
Questions about a document100.00%89.89%▼ 10.11$3.95$7.08+79 %
German
Accuracy (correct fields)Per 1,000 documents
Task · documentsGPT-6 SolMistral Large 4DifferenceGPT-6 SolMistral Large 4Difference
Invoice extraction99.95%97.49%▼ 2.46$10.65$6.72−37 %
Email triage98.22%97.85%▼ 0.37$2.73$3.78+39 %
Personal data detection99.71%99.52%▼ 0.19$2.13$2.61+23 %
Contract clauses100.00%99.44%▼ 0.56$4.36$5.12+18 %
Document summaries100.00%99.33%▼ 0.67$4.11$7.19+75 %
Questions about a document100.00%93.56%▼ 6.44$3.23$5.77+79 %
Spanish
Accuracy (correct fields)Per 1,000 documents
Task · documentsGPT-6 SolMistral Large 4DifferenceGPT-6 SolMistral Large 4Difference
Invoice extraction99.23%96.51%▼ 2.72$10.79$7.49−31 %
Email triage98.44%97.93%▼ 0.51$2.77$4.03+46 %
Personal data detection99.81%99.81%±0$2.21$3.16+43 %
Contract clauses100.00%100.00%±0$4.48$4.83+8 %
Document summaries97.89%91.33%▼ 6.56$5.02$7.99+59 %
Questions about a document100.00%94.22%▼ 5.78$3.36$5.74+71 %
French
Accuracy (correct fields)Per 1,000 documents
Task · documentsGPT-6 SolMistral Large 4DifferenceGPT-6 SolMistral Large 4Difference
Invoice extraction99.90%88.97%▼ 10.93$11.00$8.71−21 %
Email triage98.59%98.30%▼ 0.29$2.81$3.70+32 %
Personal data detection99.71%98.67%▼ 1.04$2.23$2.96+33 %
Contract clauses99.59%99.28%▼ 0.31$4.49$4.60+2 %
Document summaries97.22%72.67%▼ 24.55$4.58$11.26+146 %
Questions about a document100.00%92.33%▼ 7.67$3.43$5.74+68 %

Difference: Mistral Large 4 vs. GPT-6 Sol; accuracy in percentage points, price in percent. The better value is in bold. Click a task to see the details.

Invoice extraction · English (US) (150 documents)

GPT-6 Sol is more accurate by 4.5 pp. Mistral Large 4 is 1.5× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.94%95.39%
Error-free documents99.33%77.33%
Per 1,000 documents$9.65$6.65
To fix / 1,0007227
Speed (median)4.1 sec29.0 sec
Valid answers100.00%97.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Invoice extraction · Czech (150 documents)

GPT-6 Sol is more accurate by 2.8 pp. Mistral Large 4 is 1.8× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.90%97.08%
Error-free documents98.67%88.67%
Per 1,000 documents$10.12$5.69
To fix / 1,00013113
Speed (median)3.9 sec26.3 sec
Valid answers100.00%98.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Invoice extraction · German (150 documents)

GPT-6 Sol is more accurate by 2.5 pp. Mistral Large 4 is 1.6× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.95%97.49%
Error-free documents99.33%91.33%
Per 1,000 documents$10.65$6.72
To fix / 1,000787
Speed (median)4.9 sec35.3 sec
Valid answers100.00%98.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Invoice extraction · Spanish (150 documents)

GPT-6 Sol is more accurate by 2.7 pp. Mistral Large 4 is 1.4× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.23%96.51%
Error-free documents90.00%88.00%
Per 1,000 documents$10.79$7.49
To fix / 1,000100120
Speed (median)5.9 sec41.3 sec
Valid answers100.00%97.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Invoice extraction · French (150 documents)

GPT-6 Sol is more accurate by 10.9 pp. Mistral Large 4 is 1.3× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.90%88.97%
Error-free documents98.67%65.33%
Per 1,000 documents$11.00$8.71
To fix / 1,00013347
Speed (median)6.0 sec43.4 sec
Valid answers100.00%91.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Email triage · English (US) (150 documents)

GPT-6 Sol is more accurate by 0.3 pp. GPT-6 Sol is 1.4× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.04%98.74%
Error-free documents92.00%88.67%
Per 1,000 documents$2.44$3.52
To fix / 1,00080113
Speed (median)2.8 sec19.8 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Email triage · Czech (150 documents)

GPT-6 Sol is more accurate by 0.8 pp. GPT-6 Sol is 1.4× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields98.96%98.15%
Error-free documents90.67%89.33%
Per 1,000 documents$3.19$4.41
To fix / 1,00093107
Speed (median)2.8 sec28.4 sec
Valid answers100.00%99.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Email triage · German (150 documents)

GPT-6 Sol is more accurate by 0.4 pp. GPT-6 Sol is 1.4× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields98.22%97.85%
Error-free documents84.00%82.00%
Per 1,000 documents$2.73$3.78
To fix / 1,000160180
Speed (median)2.9 sec21.7 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Email triage · Spanish (150 documents)

GPT-6 Sol is more accurate by 0.5 pp. GPT-6 Sol is 1.5× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields98.44%97.93%
Error-free documents86.00%82.00%
Per 1,000 documents$2.77$4.03
To fix / 1,000140180
Speed (median)3.3 sec25.4 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Email triage · French (150 documents)

GPT-6 Sol is more accurate by 0.3 pp. GPT-6 Sol is 1.3× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields98.59%98.30%
Error-free documents87.33%84.67%
Per 1,000 documents$2.81$3.70
To fix / 1,000127153
Speed (median)3.2 sec24.5 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Personal data detection · English (US) (150 documents)

GPT-6 Sol is more accurate by 0.1 pp. GPT-6 Sol is 1.4× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.71%99.62%
Error-free documents98.00%97.33%
Per 1,000 documents$2.09$2.89
To fix / 1,0002027
Speed (median)2.6 sec28.2 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Personal data detection · Czech (150 documents)

GPT-6 Sol is more accurate by 0.4 pp. GPT-6 Sol is 1.4× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.71%99.33%
Error-free documents98.00%95.33%
Per 1,000 documents$2.22$3.15
To fix / 1,0002047
Speed (median)2.5 sec18.5 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Personal data detection · German (150 documents)

GPT-6 Sol is more accurate by 0.2 pp. GPT-6 Sol is 1.2× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.71%99.52%
Error-free documents98.00%96.67%
Per 1,000 documents$2.13$2.61
To fix / 1,0002033
Speed (median)2.6 sec18.6 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Personal data detection · Spanish (150 documents)

Both are equally accurate. GPT-6 Sol is 1.4× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.81%99.81%
Error-free documents98.67%98.67%
Per 1,000 documents$2.21$3.16
To fix / 1,0001313
Speed (median)2.9 sec21.9 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Personal data detection · French (150 documents)

GPT-6 Sol is more accurate by 1 pp. GPT-6 Sol is 1.3× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.71%98.67%
Error-free documents98.00%94.67%
Per 1,000 documents$2.23$2.96
To fix / 1,0002053
Speed (median)2.9 sec23.6 sec
Valid answers100.00%99.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Contract clauses · English (US) (150 documents)

GPT-6 Sol is more accurate by 3.9 pp. GPT-6 Sol is 1.6× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%96.10%
Error-free documents100.00%74.67%
Per 1,000 documents$3.33$5.24
To fix / 1,0000253
Speed (median)2.5 sec41.4 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Contract clauses · Czech (150 documents)

Both are equally accurate.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%100.00%
Error-free documents100.00%100.00%
Per 1,000 documents$4.62$4.58
To fix / 1,00000
Speed (median)2.0 sec43.9 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Contract clauses · German (150 documents)

GPT-6 Sol is more accurate by 0.6 pp. GPT-6 Sol is 1.2× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%99.44%
Error-free documents100.00%95.33%
Per 1,000 documents$4.36$5.12
To fix / 1,000047
Speed (median)2.9 sec36.2 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Contract clauses · Spanish (150 documents)

Both are equally accurate. GPT-6 Sol is 1.1× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%100.00%
Error-free documents100.00%100.00%
Per 1,000 documents$4.48$4.83
To fix / 1,00000
Speed (median)2.8 sec35.0 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Contract clauses · French (150 documents)

GPT-6 Sol is more accurate by 0.3 pp.

MetricGPT-6 SolMistral Large 4
Correct fields99.59%99.28%
Error-free documents94.67%90.67%
Per 1,000 documents$4.49$4.60
To fix / 1,0005393
Speed (median)2.6 sec34.7 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Document summaries · English (US) (150 documents)

Mistral Large 4 is more accurate by 0.4 pp. GPT-6 Sol is 1.5× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields99.56%100.00%
Error-free documents97.33%100.00%
Per 1,000 documents$3.85$5.62
To fix / 1,000270
Speed (median)8.1 sec37.2 sec
Valid answers100.00%100.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Document summaries · Czech (150 documents)

GPT-6 Sol is more accurate by 16.7 pp. GPT-6 Sol is 2.3× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%83.33%
Error-free documents100.00%83.33%
Per 1,000 documents$4.72$10.72
To fix / 1,0000167
Speed (median)7.6 sec55.4 sec
Valid answers100.00%83.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Document summaries · German (150 documents)

GPT-6 Sol is more accurate by 0.7 pp. GPT-6 Sol is 1.8× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%99.33%
Error-free documents100.00%99.33%
Per 1,000 documents$4.11$7.19
To fix / 1,00007
Speed (median)7.1 sec48.0 sec
Valid answers100.00%99.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Document summaries · Spanish (150 documents)

GPT-6 Sol is more accurate by 6.6 pp. GPT-6 Sol is 1.6× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields97.89%91.33%
Error-free documents87.33%91.33%
Per 1,000 documents$5.02$7.99
To fix / 1,00012787
Speed (median)7.0 sec38.3 sec
Valid answers100.00%91.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Document summaries · French (150 documents)

GPT-6 Sol is more accurate by 24.5 pp. GPT-6 Sol is 2.5× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields97.22%72.67%
Error-free documents83.33%72.67%
Per 1,000 documents$4.58$11.26
To fix / 1,000167273
Speed (median)6.1 sec46.4 sec
Valid answers100.00%72.67%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Questions about a document · English (US) (150 documents)

GPT-6 Sol is more accurate by 1.1 pp. GPT-6 Sol is 1.8× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%98.89%
Error-free documents100.00%97.33%
Per 1,000 documents$2.36$4.16
To fix / 1,000027
Speed (median)2.3 sec17.7 sec
Valid answers100.00%99.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Questions about a document · Czech (150 documents)

GPT-6 Sol is more accurate by 10.1 pp. GPT-6 Sol is 1.8× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%89.89%
Error-free documents100.00%78.00%
Per 1,000 documents$3.95$7.08
To fix / 1,0000220
Speed (median)2.3 sec33.6 sec
Valid answers100.00%93.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Questions about a document · German (150 documents)

GPT-6 Sol is more accurate by 6.4 pp. GPT-6 Sol is 1.8× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%93.56%
Error-free documents100.00%89.33%
Per 1,000 documents$3.23$5.77
To fix / 1,0000107
Speed (median)2.2 sec31.5 sec
Valid answers100.00%94.67%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Questions about a document · Spanish (150 documents)

GPT-6 Sol is more accurate by 5.8 pp. GPT-6 Sol is 1.7× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%94.22%
Error-free documents100.00%90.67%
Per 1,000 documents$3.36$5.74
To fix / 1,000093
Speed (median)2.5 sec22.8 sec
Valid answers100.00%95.33%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Questions about a document · French (150 documents)

GPT-6 Sol is more accurate by 7.7 pp. GPT-6 Sol is 1.7× cheaper.

MetricGPT-6 SolMistral Large 4
Correct fields100.00%92.33%
Error-free documents100.00%87.33%
Per 1,000 documents$3.43$5.74
To fix / 1,0000127
Speed (median)2.5 sec19.8 sec
Valid answers100.00%94.00%

Free with login

Accuracy field by field and the documents where the models differ.

Just your email, no password.

Log in with email

We use your email only to log you in (and for the digest if you tick it). No spam.