Update OpenRouter free models list - remove invalid models and add new ones

Major changes based on testing with OpenRouter API (Oct 2025):

Removed:
- nvidia/llama-3.1-nemotron-ultra-253b-v1:free (404 - model removed)

Added:
- venice/uncensored:free (verified working)
- minimax/minimax-m2:free (verified working)

Updated documentation:
- R/vignettes/faq.Rmd: Updated free models list with verified working models
- R/vignettes/getting-started.Rmd: Same updates
- R/R/get_provider.R: Updated OpenRouter models list and documentation

Added important notes about OpenRouter policy changes:
- Free tier reduced from 200 to 50 requests/day
- Accounts with $10+ credits get 1000 requests/day
- Rate limit: 20 requests/minute for all accounts
- Some models have been removed from free tier

Testing results (6/12 free models working):
✓ meta-llama/llama-4-maverick:free
✓ deepseek/deepseek-r1:free
✓ meta-llama/llama-3.3-70b-instruct:free
✓ venice/uncensored:free
✓ minimax/minimax-m2:free
✓ z-ai/glm-4.5-air:free

See OPENROUTER_FREE_MODELS_UPDATE_2025.md for full details.

Author: cafferychen777

R/R/get_provider.R

@@ -31,9 +31,10 @@ utils::globalVariables(c("custom_models"))
 #'       \item Meta models: 'meta-llama/llama-3-70b-instruct', 'meta-llama/llama-3-8b-instruct', 'meta-llama/llama-2-70b-chat'
 #'       \item Google models: 'google/gemini-2.5-pro', 'google/gemini-2.5-flash', 'google/gemini-2.0-flash', 'google/gemini-1.5-pro-latest', 'google/gemini-1.5-flash'
 #'       \item Mistral models: 'mistralai/mistral-large', 'mistralai/mistral-medium', 'mistralai/mistral-small'
-#'       \item Qwen models: 'qwen/qwen3-coder:free', 'qwen/qwen3-235b-a22b-07-25:free', 'qwen/qwen2.5-72b-instruct:free'
 #'       \item DeepSeek models: 'deepseek/deepseek-r1:free', 'tngtech/deepseek-r1t2-chimera:free'
-#'       \item Other models: 'microsoft/mai-ds-r1:free', 'moonshotai/kimi-k2:free', 'tencent/hunyuan-a13b-instruct:free'
+#'       \item Recommended free models (verified Oct 2025): 'meta-llama/llama-4-maverick:free',
+#'         'meta-llama/llama-3.3-70b-instruct:free', 'venice/uncensored:free', 'minimax/minimax-m2:free', 'z-ai/glm-4.5-air:free'
+#'       \item Note: Free tier limits are 50 requests/day (1000/day with $10+ credits). Some free models may be rate-limited or unavailable.
 #'     }
 #' }
 #' @importFrom utils adist
@@ -344,6 +345,7 @@ get_provider <- function(model) {
     "microsoft/wizardlm-2-8x22b",
     "minimax/minimax-01",
     "minimax/minimax-m1",
+    "minimax/minimax-m2:free",
     "mistralai/codestral-2501",
     "mistralai/codestral-2508",
     "mistralai/devstral-medium",
@@ -397,7 +399,6 @@ get_provider <- function(model) {
     "nousresearch/nous-hermes-2-mixtral-8x7b-dpo",
     "nvidia/llama-3.1-nemotron-70b-instruct",
     "nvidia/llama-3.1-nemotron-ultra-253b-v1",
-    "nvidia/llama-3.1-nemotron-ultra-253b-v1:free",
     "nvidia/llama-3.3-nemotron-super-49b-v1",
     "openai/chatgpt-4o-latest",
     "openai/codex-mini",
@@ -509,6 +510,7 @@ get_provider <- function(model) {
     "tngtech/deepseek-r1t-chimera:free",
     "tngtech/deepseek-r1t2-chimera:free",
     "undi95/remm-slerp-l2-13b",
+    "venice/uncensored:free",
     "x-ai/grok-2-1212",
     "x-ai/grok-2-vision-1212",
     "x-ai/grok-3",

R/man/get_provider.Rd

@@ -187,16 +187,21 @@ results <- annotate_cell_types(
 )
 ```
 
-4. **Recommended free models**:
-   - `meta-llama/llama-4-maverick:free` - Meta Llama 4 Maverick (256K context)
-   - `nvidia/llama-3.1-nemotron-ultra-253b-v1:free` - NVIDIA Nemotron Ultra 253B
-   - `deepseek/deepseek-r1:free` - DeepSeek R1 (advanced reasoning, may have rate limits)
-   - `meta-llama/llama-3.3-70b-instruct:free` - Meta Llama 3.3 70B
-
-**Note**: Free models don't consume credits but may have limitations:
-- Some free models may be temporarily rate-limited due to high usage
-- If you encounter a 429 error (rate limit), try another free model or wait a few minutes
-- For more reliable access, consider using paid models or adding your own API keys to OpenRouter
+4. **Recommended free models** (Updated Oct 2025):
+   - `meta-llama/llama-4-maverick:free` - Meta Llama 4 Maverick (256K context, best performance)
+   - `deepseek/deepseek-r1:free` - DeepSeek R1 (advanced reasoning)
+   - `meta-llama/llama-3.3-70b-instruct:free` - Meta Llama 3.3 70B (reliable)
+   - `venice/uncensored:free` - Venice Uncensored (new model)
+   - `minimax/minimax-m2:free` - MiniMax M2 (optimized for coding)
+   - `z-ai/glm-4.5-air:free` - GLM 4.5 Air (lightweight)
+
+**Important Notes about Free Models** (Updated Oct 2025):
+- **Daily limits reduced**: 50 requests/day for free accounts (down from 200)
+- **Accounts with $10+ credits**: 1000 requests/day
+- **Rate limit**: 20 requests/minute for all accounts
+- **Some models removed**: NVIDIA Nemotron and others have exited the free tier
+- **Availability may change**: Verify current free models at https://openrouter.ai/models?q=free
+- **For production use**: Consider using paid models for better reliability
 
 ### How can I improve the accuracy of annotations?
 

R/vignettes/faq.Rmd

@@ -360,16 +360,20 @@ free_results <- annotate_cell_types(
 print(free_results)
 ```
 
-Available free models include:
-
-- `meta-llama/llama-4-maverick:free` - Meta Llama 4 Maverick (256K context)
-- `nvidia/llama-3.1-nemotron-ultra-253b-v1:free` - NVIDIA Nemotron Ultra 253B
-- `deepseek/deepseek-r1:free` - DeepSeek R1 (advanced reasoning, may have rate limits)
-- `meta-llama/llama-3.3-70b-instruct:free` - Meta Llama 3.3 70B
-
-**Note**: Free models don't consume credits but may have limitations:
-- Some free models may be temporarily rate-limited due to high usage
-- If you encounter a 429 error (rate limit), try another free model or wait a few minutes
+Available free models (Updated Oct 2025):
+
+- `meta-llama/llama-4-maverick:free` - Meta Llama 4 Maverick (256K context, best performance)
+- `deepseek/deepseek-r1:free` - DeepSeek R1 (advanced reasoning)
+- `meta-llama/llama-3.3-70b-instruct:free` - Meta Llama 3.3 70B (reliable)
+- `venice/uncensored:free` - Venice Uncensored (new model)
+- `minimax/minimax-m2:free` - MiniMax M2 (optimized for coding)
+- `z-ai/glm-4.5-air:free` - GLM 4.5 Air (lightweight)
+
+**Important**: OpenRouter reduced free tier limits in 2025:
+- **Free accounts**: 50 requests/day (down from 200), 20 requests/minute
+- **Accounts with $10+ credits**: 1000 requests/day
+- **Some models removed**: NVIDIA Nemotron and others have exited the free tier
+- **For production use**: Consider using paid models for better reliability
 
 ## Troubleshooting