LLM API price changes
Every change PromptSpend has recorded to an LLM API price, newest first. Since tracking began on 2026-08-02, vendors have changed a published rate 42 times on 9 days, 62 models were added and 0 were removed.
How to read it. A price change is the vendor changing a rate. A correction is PromptSpend reading a better page, so the figure changes here but the vendor’s price did not; it is labeled apart so repricing is never overstated. A rate PromptSpend only began recording is not a change to anyone’s bill and is left out.
The same changes, one entry a day: the Atom feed. Every row, with the commit that published it: the price-history dataset. Current prices with their sources are on the models page.
2026-11-17
Models added
- GPT 6.2 Astra — $8 in, $40 out per 1M tokens
- GPT 6.2 Sol — $2 in, $10 out per 1M tokens
2026-11-12
Models added
- Gemini 4 Hydrogen — $1.5 in, $7.5 out per 1M tokens
2026-11-11
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| Mistral Large 4 | Input | $1.36 | $0.7 | −49% cut |
| Mistral Large 4 | Output | $4.18 | $2.1 | −50% cut |
| Mistral Large 4 | Cached input | $0.14 | $0.07 | −50% cut |
| Mistral Large 4.0 | Input | $1.36 | $0.7 | −49% cut |
| Mistral Large 4.0 | Output | $4.18 | $2.1 | −50% cut |
| Mistral Large 4.0 | Cached input | $0.14 | $0.07 | −50% cut |
Corrections
PromptSpend had been reading the wrong page. The vendor’s rate did not move.
| Model | Rate | Was | Now |
|---|---|---|---|
| Gemini 4 Carbon | Input | $2 | $4 |
| Gemini 4 Carbon | Output | $10 | $20 |
| Gemini 4 Carbon | Cached input | $0.1 | $0.2 |
2026-11-10
Models added
- Gemini 4 Carbon — $2 in, $10 out per 1M tokens
2026-11-09
Models added
- Claude Fable 5 2 — $8 in, $40 out per 1M tokens
2026-10-29
Models added
- Nova Lite V1:0 — $0.06 in, $0.24 out per 1M tokens
- Nova Micro V1:0 — $0.035 in, $0.14 out per 1M tokens
- Nova Pro V1:0 — $0.8 in, $3.2 out per 1M tokens
- Command R 08 2024 — $0.15 in, $0.6 out per 1M tokens
- Command R Plus 08 2024 — $2.5 in, $10 out per 1M tokens
- Qwen3.8 Flash — $0.15 in, $0.47 out per 1M tokens
- GPT 5 Pro — $15 in, $120 out per 1M tokens
- GPT 5.2 Pro — $21 in, $168 out per 1M tokens
- GPT 5.4 Pro — $30 in, $180 out per 1M tokens
- GPT 5.5 Pro — $30 in, $180 out per 1M tokens
- Codestral 2508 — $0.3 in, $0.9 out per 1M tokens
- Kimi K2.7 Code — $0.95 in, $4 out per 1M tokens
- Kimi K3.1 Pro — $2 in, $10 out per 1M tokens
- O1 Pro — $150 in, $600 out per 1M tokens
- O3 Pro — $20 in, $80 out per 1M tokens
- Qwen3.8 Omni Flash — $0.15 in, $0.47 out per 1M tokens
2026-10-27
Models added
- GPT 6.1 Terra — $1 in, $5 out per 1M tokens
2026-10-24
Models added
- Qwen 3.8 Max — $2 in, $6 out per 1M tokens
- Gemini 2.5 Flash-Lite — $0.1 in, $0.4 out per 1M tokens
- Gemini 4 Argon — $4 in, $20 out per 1M tokens
- GPT-4.1 — $2 in, $8 out per 1M tokens
- GPT-4.1 Mini — $0.4 in, $1.6 out per 1M tokens
- GPT-4o — $2.5 in, $10 out per 1M tokens
- GPT-4o Mini — $0.15 in, $0.6 out per 1M tokens
- GPT-6 Astra — $10 in, $50 out per 1M tokens
- GPT-6 Luna — $0.1 in, $0.5 out per 1M tokens
- GPT-6 Sol — $2 in, $10 out per 1M tokens
- GPT-6.1 Sol — $2 in, $10 out per 1M tokens
- Grok Build 0.1 — $1 in, $2 out per 1M tokens
- GLM-4 32B — $0.1 in, $0.1 out per 1M tokens
- GLM-4.5 — $0.6 in, $2.2 out per 1M tokens
- GLM-4.5-Air — $0.2 in, $1.1 out per 1M tokens
- GLM-4.5-AirX — $1.1 in, $4.5 out per 1M tokens
- GLM-4.5-X — $2.2 in, $8.9 out per 1M tokens
- GLM-5.3 Flash — $0.15 in, $0.5 out per 1M tokens
- MAI-Thinking-1 — $2 in, $8 out per 1M tokens
- Mercury 2 — $0.25 in, $0.75 out per 1M tokens
- Mercury 2.5 — $0.2 in, $0.75 out per 1M tokens
- Muse Spark 1.3 — $1.25 in, $4.25 out per 1M tokens
- Muse Wonder 1.0 — $4 in, $20 out per 1M tokens
- MiMo V2.6 Flash — $0.14 in, $0.28 out per 1M tokens
- MiMo V2.6 Pro — $0.435 in, $0.87 out per 1M tokens
2026-10-21
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| Claude Sonnet 5.5 | Cached input | $0.2 | $0.1 | −50% cut |
Corrections
PromptSpend had been reading the wrong page. The vendor’s rate did not move.
| Model | Rate | Was | Now |
|---|---|---|---|
| Mistral Large 4 | Input | $0.68 | $1.36 |
| Mistral Large 4 | Output | $2.09 | $4.18 |
| Mistral Large 4 | Cached input | $0.068 | $0.14 |
| Mistral Large 4.0 | Input | $0.68 | $1.36 |
| Mistral Large 4.0 | Output | $2.09 | $4.18 |
| Mistral Large 4.0 | Cached input | $0.068 | $0.14 |
2026-10-08
Models added
- Claude Haiku 5 5 — $0.1 in, $0.5 out per 1M tokens
2026-10-07
Models added
- Mistral Large 4 — $0.68 in, $2.09 out per 1M tokens
- Mistral Large 4 0 — $0.68 in, $2.09 out per 1M tokens
2026-10-03
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| Nova 2 Pro | Input | $2.1875 | $1.25 | −43% cut |
| Nova 2 Pro | Output | $17.5 | $10 | −43% cut |
| Nova 2 Pro | Cached input | $0.5469 | $0.3125 | −43% cut |
2026-09-29
Models added
- Claude Sonnet 5 5 — $2 in, $10 out per 1M tokens
2026-09-23
Models added
- Claude Opus 5 5 — $4 in, $20 out per 1M tokens
2026-09-22
Models added
- Grok 4.7 — $2 in, $6 out per 1M tokens
2026-09-16
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| DeepSeek V4 Flash | Input | $0.44 | $0.3 | −32% cut |
| DeepSeek V4 Flash | Output | $1.32 | $1.2 | −9% cut |
| DeepSeek V4 Flash | Cached input | $0.014 | $0.006 | −57% cut |
2026-09-04
Models added
- Gemini 3.8 Flash — $0.75 in, $3.75 out per 1M tokens
2026-09-02
Models added
- Claude Fable 5 1 — $10 in, $50 out per 1M tokens
- GLM 5.2 — $1.4 in, $4.4 out per 1M tokens
2026-09-01
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| Claude Sonnet 5 | Input | $3 | $2 | −33% cut |
| Claude Sonnet 5 | Output | $15 | $10 | −33% cut |
| Claude Sonnet 5 | Cached input | $0.3 | $0.2 | −33% cut |
| Claude Sonnet 5 | Cache write | $3.75 | $2.5 | −33% cut |
2026-08-30
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| Grok 3 | Input | $3 | $1.25 | −58% cut |
| Grok 3 | Output | $15 | $2.5 | −83% cut |
| Grok 3 | Cached input | $0.75 | $0.2 | −73% cut |
| Grok 4 | Input | $3 | $1.25 | −58% cut |
| Grok 4 | Output | $15 | $2.5 | −83% cut |
| Grok 4.1 Fast | Input | $0.2 | $1.25 | +525% rise |
| Grok 4.1 Fast | Output | $0.5 | $2.5 | +400% rise |
| Grok 4.1 Fast | Cached input | $0.05 | $0.2 | +300% rise |
Models added
- Grok 4.20 — $1.25 in, $2.5 out per 1M tokens
2026-08-28
Models added
- Mistral Medium 3 — $1.5 in, $7.5 out per 1M tokens
- GLM 5.3 — $1.4 in, $4.4 out per 1M tokens
2026-08-23
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| GPT-5.6 (routes to Sol) | Input | $5 | $4 | −20% cut |
| GPT-5.6 (routes to Sol) | Output | $30 | $20 | −33% cut |
| GPT-5.6 (routes to Sol) | Cached input | $0.5 | $0.4 | −20% cut |
2026-08-22
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| Claude Sonnet 5 | Input | $2 | $3 | +50% rise |
| Claude Sonnet 5 | Output | $10 | $15 | +50% rise |
| Claude Sonnet 5 | Cached input | $0.2 | $0.3 | +50% rise |
| Claude Sonnet 5 | Cache write | $2.5 | $3.75 | +50% rise |
Models added
- Kimi K3 — $3 in, $15 out per 1M tokens
2026-08-16
Price changes
The vendor changed a published rate.
| Model | Rate | Was | Now | Change |
|---|---|---|---|---|
| Claude Sonnet 5 | Input | $3 | $2 | −33% cut |
| Claude Sonnet 5 | Output | $15 | $10 | −33% cut |
| Claude Sonnet 5 | Cached input | $0.3 | $0.2 | −33% cut |
| Claude Sonnet 5 | Cache write | $3.75 | $2.5 | −33% cut |
| DeepSeek V4 Flash | Input | $0.14 | $0.44 | +214% rise |
| DeepSeek V4 Flash | Output | $0.28 | $1.32 | +371% rise |
| DeepSeek V4 Flash | Cached input | $0.0028 | $0.014 | +400% rise |
| DeepSeek V4 Pro | Input | $0.435 | $1.32 | +203% rise |
| DeepSeek V4 Pro | Output | $0.87 | $3.96 | +355% rise |
| DeepSeek V4 Pro | Cached input | $0.0036 | $0.044 | +1114% rise |
Models added
- Gemini 3.7 Flash — $1.5 in, $7.5 out per 1M tokens
- Grok 4.6 — $2 in, $6 out per 1M tokens
2026-08-06
Corrections
PromptSpend had been reading the wrong page. The vendor’s rate did not move.
| Model | Rate | Was | Now |
|---|---|---|---|
| Grok 4.5 | Cached input | $0.5 | $0.3 |
2026-08-02
PromptSpend began tracking on this day, with 70 models in the catalog. Nothing before it is on record here.
Common questions
How often do LLM API prices change?
Since PromptSpend began tracking on 2026-08-02, vendors have changed a published rate 42 times, spread over 9 days. 25 days have had any news at all, counting new and removed models. Earlier changes are not on record here.
What counts as a price change on this page?
A price change is a vendor changing a rate it publishes. A correction is PromptSpend reading a better page, so the figure here changes but the vendor's price did not; corrections are labeled apart so repricing is never overstated. A rate PromptSpend only began recording is not a change to anyone's bill and is left out.
Can I get these price changes as a feed or a file?
Yes. The same changes are published as an Atom feed with one entry a day, and every row, with the commit that published it, is in the price-history dataset as CSV and JSON. Neither needs a key or an account.