Tokens served
241.1B
Every token FreeTheAI has served since it opened, across 8.2M requests from people in 145 countries.
- Requests served
- 8.2M
- Saved by users
- $53,567
- Active, 30 days
- 5,634
- Models live
- 296
Daily usage
50.7B tokens and 1.4M requests in the last 30 days.
| Tokens | Requests | |
|---|---|---|
| Oct 10 | 1.6B | 41.9K |
| Oct 11 | 1.6B | 45.3K |
| Oct 12 | 1.7B | 47K |
| Oct 13 | 1.6B | 45.3K |
| Oct 14 | 1.7B | 45.3K |
| Oct 15 | 1.5B | 46K |
| Oct 16 | 1.7B | 46.8K |
| Oct 17 | 1.7B | 45.8K |
| Oct 18 | 2B | 49.1K |
| Oct 19 | 1.6B | 52.9K |
| Oct 20 | 1.4B | 43.6K |
| Oct 21 | 1.4B | 41.3K |
| Oct 22 | 1.5B | 41.2K |
| Oct 23 | 1.5B | 42K |
| Oct 24 | 1.6B | 43.6K |
| Oct 25 | 1.8B | 47K |
| Oct 26 | 1.8B | 47.7K |
| Oct 27 | 1.6B | 44.3K |
| Oct 28 | 1.5B | 44K |
| Oct 29 | 1.5B | 44.1K |
| Oct 30 | 1.8B | 48.5K |
| Oct 31 | 1.7B | 49.7K |
| Nov 1 | 2B | 55K |
| Nov 2 | 2B | 54.1K |
| Nov 3 | 1.9B | 51.4K |
| Nov 4 | 2B | 52.5K |
| Nov 5 | 1.8B | 49.6K |
| Nov 6 | 2B | 54K |
| Nov 7 | 2.1B | 56.2K |
| Nov 8 | 1.1B | 28.6K |
All time: 229.3B input tokens (179.1B of them cached) and 11.7B output tokens. Daily numbers cover FreeTheAI traffic. All-time totals also include the original free API.
Top models
By tokens served in the last 7 days.
Tokens per day by model, last 30 days
- GLM-5.334.5B
- Kimi K2.7 Code3.9B
- GLM-5.3 Flash3.6B
- MiniMax M32.1B
- Gemma4 31B1.7B
- K31.2B
- Gemma4 31B151.1M
- Nemotron 3 Ultra 550B A55B139.1M
- Others
| Day | GLM-5.3 | Kimi K2.7 Code | GLM-5.3 Flash | MiniMax M3 | Gemma4 31B | K3 | Gemma4 31B | Nemotron 3 Ultra 550B A55B | Others | Total |
|---|---|---|---|---|---|---|---|---|---|---|
| Oct 10 | 983.5M | 55.3M | 134.9M | 66.6M | 48.7M | 58.6M | 5.3K | 4.2M | 228M | 1.6B |
| Oct 11 | 977M | 78.8M | 149.5M | 81.9M | 39.5M | 2.2M | 9.7M | 4.7M | 246.2M | 1.6B |
| Oct 12 | 1.1B | 73.8M | 80M | 91.9M | 33.7M | 0 | 17.1M | 2.1M | 276.7M | 1.7B |
| Oct 13 | 1.1B | 73M | 94.3M | 96.2M | 30.5M | 21.3M | 3.6M | 2.7M | 221.4M | 1.6B |
| Oct 14 | 1.1B | 71.3M | 83M | 76.9M | 17.2M | 49.1M | 5.1M | 3M | 206.2M | 1.7B |
| Oct 15 | 940.4M | 96.9M | 116.2M | 90.9M | 30.3M | 46.2M | 3.5M | 5.1M | 214.3M | 1.5B |
| Oct 16 | 1.1B | 75.6M | 93.2M | 101.2M | 44.3M | 52.5M | 3.1M | 21.3K | 232.4M | 1.7B |
| Oct 17 | 1.1B | 82.3M | 83.7M | 79.6M | 30.3M | 72M | 4M | 98.3K | 213.2M | 1.7B |
| Oct 18 | 1.3B | 103.6M | 132.1M | 96.8M | 41.6M | 109M | 9.6M | 11.9M | 226.9M | 2B |
| Oct 19 | 1B | 104.3M | 78M | 100.4M | 26.3M | 82.3M | 4.6M | 3.4M | 246.7M | 1.6B |
| Oct 20 | 894.5M | 94.8M | 129.9M | 71.1M | 45.6M | 64M | 2.8M | 2.3M | 122.4M | 1.4B |
| Oct 21 | 933.3M | 100.9M | 117.9M | 69.8M | 37.6M | 52.1M | 710.6K | 1.6M | 90M | 1.4B |
| Oct 22 | 1B | 101.3M | 92.2M | 50.4M | 71.1M | 47M | 1.4M | 1.5M | 69M | 1.5B |
| Oct 23 | 1.1B | 111.2M | 146.5M | 31.9M | 53M | 15.1M | 3.6M | 2.4M | 46.5M | 1.5B |
| Oct 24 | 1.3B | 119.7M | 122.2M | 40M | 45.6M | 13.3M | 2.5K | 2.6M | 32.5M | 1.6B |
| Oct 25 | 1.3B | 112.2M | 215.8M | 57.3M | 47.7M | 37.9M | 0 | 3.2M | 45.1M | 1.8B |
| Oct 26 | 1.3B | 94M | 141M | 45.1M | 51.7M | 114.6M | 1.4M | 2.3M | 48.6M | 1.8B |
| Oct 27 | 1.1B | 122.2M | 85.8M | 114.4M | 55.8M | 71.4M | 4M | 2.4M | 84M | 1.6B |
| Oct 28 | 1B | 145M | 116.1M | 31.6M | 49.1M | 52.9M | 7.9M | 5.2M | 68.5M | 1.5B |
| Oct 29 | 1.1B | 145.7M | 139.7M | 35.9M | 49.7M | 29.2M | 1.3M | 5.7M | 54.4M | 1.5B |
| Oct 30 | 1.2B | 137.4M | 188.1M | 58.4M | 51.2M | 27M | 12.2M | 2.6M | 60.8M | 1.8B |
| Oct 31 | 1.2B | 164.7M | 115.7M | 50.4M | 74M | 25.3M | 7.6M | 9.2M | 59.7M | 1.7B |
| Nov 1 | 1.4B | 163.2M | 165.5M | 73.6M | 106.2M | 22M | 7.7M | 6.9M | 56.2M | 2B |
| Nov 2 | 1.4B | 207.7M | 122.1M | 76.7M | 80.9M | 26M | 7.4M | 5M | 44.6M | 2B |
| Nov 3 | 1.4B | 175.8M | 129M | 80.6M | 56.5M | 32.5M | 5.5M | 7.3M | 48.3M | 1.9B |
| Nov 4 | 1.4B | 203.9M | 116.8M | 101.7M | 87.9M | 21.8M | 7M | 5.5M | 57.8M | 2B |
| Nov 5 | 1.1B | 291.8M | 87.8M | 100.6M | 94.7M | 25.8M | 5.2M | 12.3M | 42.8M | 1.8B |
| Nov 6 | 1.4B | 221.1M | 110.9M | 84.3M | 106.5M | 25.2M | 10.1M | 8.3M | 35.1M | 2B |
| Nov 7 | 1.6B | 231.6M | 122.7M | 59.2M | 101.6M | 16.4M | 3.8M | 11.6M | 21.7M | 2.1B |
| Nov 8 | 839.5M | 121.9M | 87.6M | 24.8M | 44M | 7.7M | 1.1M | 3.8M | 7.3M | 1.1B |
- 1GLM-5.3by Zhipu AI10.4B70%up 16%
- 2Kimi K2.7 Codeby Moonshot1.6B11%up 62%
- 3GLM-5.3 Flashby Zhipu AI941.7M6.4%down 15%
- 4Gemma4 31Bby Google678.1M4.6%up 67%
- 5MiniMax M3by MiniMax601.6M4.1%up 47%
- 6K3by Moonshot177.4M1.2%down 53%
- 7Nemotron 3 Ultra 550B A55Bby NVIDIA60.6M0.4%up 85%
- 8DeepSeek Chatby DeepSeek51.3M0.3%up 18%
- 9Gemma4 31Bby Google47.9M0.3%up 36%
- 10MiniMax M2.7by MiniMax28.1M0.2%up 300%
People per day by model
Accounts that used each model each day, last 30 days. The legend shows each model's daily average.
- GLM-5.31K
- K3267
- Kimi K2.7 Code207
- GLM-5.3 Flash193
- MiniMax M3142
- Gemma4 31B103
- DeepSeek Chat44
- Nemotron 3 Ultra 550B A55B32
- Others
| Day | GLM-5.3 | K3 | Kimi K2.7 Code | GLM-5.3 Flash | MiniMax M3 | Gemma4 31B | DeepSeek Chat | Nemotron 3 Ultra 550B A55B | Others | Total |
|---|---|---|---|---|---|---|---|---|---|---|
| Oct 10 | 821 | 310 | 126 | 148 | 141 | 55 | 40 | 24 | 1.4K | 3K |
| Oct 11 | 855 | 298 | 124 | 155 | 148 | 50 | 38 | 19 | 1.4K | 3K |
| Oct 12 | 873 | 268 | 129 | 156 | 152 | 53 | 32 | 24 | 1.3K | 2.9K |
| Oct 13 | 890 | 269 | 147 | 170 | 151 | 52 | 37 | 22 | 1.3K | 3.1K |
| Oct 14 | 890 | 247 | 144 | 151 | 146 | 50 | 34 | 22 | 1.2K | 2.8K |
| Oct 15 | 905 | 262 | 150 | 160 | 144 | 58 | 30 | 23 | 1.1K | 2.9K |
| Oct 16 | 911 | 260 | 136 | 161 | 165 | 62 | 36 | 30 | 1.2K | 3K |
| Oct 17 | 918 | 242 | 133 | 164 | 155 | 53 | 21 | 26 | 1.2K | 2.9K |
| Oct 18 | 889 | 269 | 137 | 164 | 151 | 50 | 32 | 26 | 1.3K | 3K |
| Oct 19 | 885 | 233 | 144 | 153 | 143 | 52 | 34 | 20 | 1.5K | 3.2K |
| Oct 20 | 902 | 257 | 154 | 150 | 158 | 56 | 25 | 15 | 1.2K | 2.9K |
| Oct 21 | 924 | 279 | 173 | 180 | 155 | 64 | 32 | 26 | 1.5K | 3.4K |
| Oct 22 | 971 | 302 | 186 | 173 | 150 | 84 | 43 | 28 | 1.7K | 3.7K |
| Oct 23 | 966 | 244 | 190 | 190 | 141 | 99 | 38 | 29 | 1.5K | 3.4K |
| Oct 24 | 1K | 231 | 213 | 184 | 137 | 99 | 35 | 36 | 1.5K | 3.5K |
| Oct 25 | 996 | 247 | 188 | 192 | 118 | 97 | 36 | 27 | 1.4K | 3.3K |
| Oct 26 | 1K | 283 | 192 | 183 | 115 | 99 | 38 | 22 | 1.7K | 3.7K |
| Oct 27 | 1K | 342 | 237 | 251 | 166 | 128 | 43 | 29 | 1.9K | 4.1K |
| Oct 28 | 1.1K | 313 | 235 | 243 | 149 | 129 | 44 | 45 | 2.1K | 4.3K |
| Oct 29 | 1.1K | 316 | 242 | 265 | 132 | 133 | 47 | 41 | 1.8K | 4.1K |
| Oct 30 | 1.1K | 261 | 255 | 197 | 133 | 125 | 52 | 38 | 1.9K | 4.1K |
| Oct 31 | 1.2K | 305 | 277 | 253 | 159 | 160 | 60 | 47 | 3.4K | 5.8K |
| Nov 1 | 1.2K | 301 | 272 | 244 | 135 | 171 | 59 | 54 | 3.5K | 6K |
| Nov 2 | 1.2K | 268 | 285 | 234 | 146 | 161 | 44 | 51 | 3K | 5.4K |
| Nov 3 | 1.2K | 261 | 263 | 220 | 116 | 146 | 40 | 38 | 2.1K | 4.4K |
| Nov 4 | 1.2K | 252 | 294 | 226 | 144 | 164 | 71 | 31 | 2.6K | 5K |
| Nov 5 | 1.2K | 283 | 352 | 269 | 159 | 197 | 88 | 54 | 2.4K | 5K |
| Nov 6 | 1.2K | 228 | 315 | 224 | 138 | 167 | 67 | 43 | 2.7K | 5.1K |
| Nov 7 | 1.2K | 234 | 301 | 203 | 131 | 165 | 71 | 43 | 2.7K | 5.1K |
| Nov 8 | 932 | 130 | 213 | 141 | 78 | 124 | 48 | 31 | 1.9K | 3.6K |
Weekly retention
Of the people who used a model the week before last, the share who used it again last week.
- GLM-5.31,423 of 2,085 came back68%
- Kimi K2.7 Code348 of 577 came back60%
- Gemma4 31B199 of 375 came back53%
- GLM-5.3 Flash304 of 681 came back45%
- K3424 of 977 came back43%
- MiniMax M3190 of 504 came back38%
- DeepSeek Chat60 of 182 came back33%
- Nemotron 3 Ultra 550B A55B40 of 154 came back26%
- GPT-6 Luna33 of 143 came back23%
- Claude Opus 5.530 of 157 came back19%
Cache hits
Share of input tokens served from the prompt cache, last 7 days.
- Nemotron 3.5 Lightning23.1M input tokens88%
- MiniMax M3511.2M input tokens74%
- Kimi K2.7 Code1.4B input tokens69%
- GLM-5.38.5B input tokens54%
- Nemotron 3 Ultra 550B A55B51.6M input tokens50%
- Gemma4 31B556.4M input tokens47%
- K3148.5M input tokens46%
- GLM-5.3 Flash729M input tokens41%
- DeepSeek Chat43.2M input tokens0.0%
- Gemma4 31B39.3M input tokens0.0%
Price per 1M tokens
What the most used models cost here, beside their maker's own price.
- GLM-5.3Maker price $1.40 in, $4.40 outFree
- Kimi K2.7 CodeMaker price $0.95 in, $4.00 outFree
- GLM-5.3 FlashMaker price $0.15 in, $0.50 outFree
- MiniMax M3Maker price $0.30 in, $1.20 outFree
- Gemma4 31BMaker price $0.40 in, $1.00 outFree
- K3Maker price $3.00 in, $15.00 outFree
- Gemma4 31BMaker price $0.40 in, $1.00 outFree
- Nemotron 3 Ultra 550B A55BMaker price $0.60 in, $2.50 outFree
Market share by maker
How the top makers compare in the last 7 days: share, success, and speed.
| Token share | Request share | Success | Speed | Quick replies | |
|---|---|---|---|---|---|
| Zhipu AI | 77% | 75% | 58.1% | 54 tok/s | 41 s |
| Moonshot | 12% | 12% | 81.8% | 97 tok/s | 19 s |
| 4.9% | 4.9% | 98.6% | 185 tok/s | 6.5 s | |
| MiniMax | 4.3% | 2.6% | 99.3% | 93 tok/s | 10 s |
| NVIDIA | 0.8% | 1.4% | 94.6% | 72 tok/s | 19 s |
- Zhipu AI77%
- Moonshot12%
- Google4.9%
- MiniMax4.3%
- NVIDIA0.8%
- Others1.1%
Top clients
By requests, all time, including the original free API.
| Requests | Tokens per request | Reply length | Cache hits | Models used | |
|---|---|---|---|---|---|
| janitor | 2.8M | 23.9K | 1.7K | 51% | 58 |
| node | 1.2M | 53.7K | 2.6K | 46% | 134 |
| openai-sdk | 511.5K | 88.8K | 1.1K | 84% | 114 |
| bun | 358.1K | 85.4K | 4.4K | 52% | 99 |
| tavo | 270.3K | 33.3K | 2.6K | 36% | 87 |
| okhttp | 232.2K | 27.8K | 1.5K | 48% | 82 |
- janitor2.8M
- node1.2M
- openai-sdk511.5K
- bun358.1K
- tavo270.3K
- okhttp232.2K
Where people use FreeTheAI
Accounts by the country of their latest request in the last 30 days.
- Accounts
- 8,635
- Countries
- 145
Top countries
- United States855
- Indonesia328
- India313
- Brazil272
- Germany272
- Philippines271
- China232
- Netherlands219
- Vietnam138
- Russia130
2,604 more users in 135 other countries.
Countries with fewer than 3 users are counted together. Updated every 30 minutes.
Keeping the free tier fair
1,714 accounts are stopped for breaking the free-tier rules, 1,228 of them in the last 30 days.
The rules: one account per person, no sharing or reselling access, and no extra accounts to get more free use. Think we got it wrong? Send an appeal from your dashboard.
How we count
- Tokens
- Input tokens (cached ones included) plus output tokens, as the provider reported them for each request on FreeTheAI. Days run on UTC, and today is still filling in. Only models in the public catalog are named.
- People
- A person is one FreeTheAI account. Someone who used two models on one day counts once for each. Retention compares the week before last with last week and lists only models with at least 20 people in the first week.
- Cache hits
- Input tokens the provider read from its prompt cache, as a share of all input tokens over the last 7 days. A higher share means repeated context costs less and comes back faster.
- Prices
- What FreeTheAI charges per million input and output tokens. The maker price is the model maker's own published price, shown for comparison.






