Listen on Youtube Music, Spotify and other streaming platforms. A new tournament will be released every month: follow me on twitter to not miss anything.
Are you frustrated by AI companies training on benchmarks? Do you enjoy battle rap? Well, both of you are at the right place. Rap battles combine debate, poetry and improvisation - the three toughest tests of verbal intelligence. State of the art in LLM evaluation. State of the art in generative poetry. The benchmark to end all benchmarks. Built with keeptalking.
Round 7
Round 6
Round 5
- qwen/qwen3-30b-a3b-thinking-2507 v openai/o3-pro lyrics, verdicts
- openai/gpt-4-1106-preview v qwen/qwen3-235b-a22b-2507 lyrics, verdicts
- qwen/qwen3-235b-a22b-thinking-2507 v qwen/qwen-plus lyrics, verdicts
Round 4
- qwen/qwen3-max v openai/gpt-4-1106-preview lyrics, verdicts
- openai/o4-mini v qwen/qwen3-235b-a22b-thinking-2507 lyrics, verdicts
- deepseek/deepseek-v3.1-terminus v openai/o3-pro lyrics, verdicts
- mistralai/mistral-medium-3.1 v qwen/qwen3-235b-a22b-2507 lyrics, verdicts
- qwen/qwen-plus v qwen/qwen-max lyrics, verdicts
Round 3
- openai/gpt-4o-2024-05-13 v qwen/qwen3-30b-a3b-thinking-2507 lyrics, verdicts
- qwen/qwen3-max v google/gemini-2.5-flash-lite-preview-06-17 lyrics, verdicts
- deepseek/deepseek-r1 v deepseek/deepseek-v3.1-terminus lyrics, verdicts
- mistralai/mistral-medium-3.1 v qwen/qwen3-32b lyrics, verdicts
- bytedance/seed-oss-36b-instruct v qwen/qwen-plus lyrics, verdicts
- qwen/qwen3-14b v openai/gpt-4-1106-preview lyrics, verdicts
- qwen/qwen3-235b-a22b-thinking-2507 v openai/gpt-5 lyrics, verdicts
- openai/o3-pro v anthropic/claude-3-opus lyrics, verdicts
- qwen/qwen3-235b-a22b-2507 v anthropic/claude-3.5-sonnet lyrics, verdicts
- nvidia/llama-3.1-nemotron-ultra-253b-v1 v qwen/qwen-max lyrics, verdicts
- openai/chatgpt-4o-latest v openai/o4-mini lyrics, verdicts
Round 2
- anthropic/claude-sonnet-4 v qwen/qwen3-max lyrics, verdicts
- deepseek/deepseek-r1 v meta-llama/llama-3.1-405b-instruct lyrics, verdicts
- microsoft/wizardlm-2-8x22b v deepseek/deepseek-v3.1-terminus lyrics, verdicts
- microsoft/mai-ds-r1 v qwen/qwen-plus lyrics, verdicts
- qwen/qwq-32b v openai/gpt-4-1106-preview lyrics, verdicts
- openai/gpt-5 v perplexity/sonar-reasoning-pro lyrics, verdicts
- qwen/qwen3-235b-a22b v anthropic/claude-3-opus lyrics, verdicts
- google/gemini-2.5-flash-lite v anthropic/claude-3.5-sonnet lyrics, verdicts
- qwen/qwen3-30b-a3b v qwen/qwen-max lyrics, verdicts
- openai/o4-mini v deepseek/deepseek-chat lyrics, verdicts
- openai/gpt-oss-20b v mistralai/mistral-medium-3.1 lyrics, verdicts
- bytedance/seed-oss-36b-instruct v stepfun-ai/step3 lyrics, verdicts
- qwen/qwen3-14b v arcee-ai/virtuoso-large lyrics, verdicts
- meituan/longcat-flash-chat v qwen/qwen3-235b-a22b-thinking-2507 lyrics, verdicts
- openai/o3-pro v openai/o4-mini-high lyrics, verdicts
- qwen/qwen3-235b-a22b-2507 v alibaba/tongyi-deepresearch-30b-a3b lyrics, verdicts
- nvidia/llama-3.1-nemotron-ultra-253b-v1 v tngtech/deepseek-r1t-chimera lyrics, verdicts
- openai/gpt-4-turbo v openai/chatgpt-4o-latest lyrics, verdicts
- anthropic/claude-3.5-sonnet-20240620 v qwen/qwen3-30b-a3b-thinking-2507 lyrics, verdicts
- google/gemini-2.5-flash-lite-preview-06-17 v deepseek/deepseek-v3.1-base lyrics, verdicts
- google/gemma-3-27b-it v qwen/qwen3-32b lyrics, verdicts
Round 1
- x-ai/grok-4-fast v openai/gpt-4o-2024-05-13 lyrics, verdicts
- anthropic/claude-sonnet-4 v mistralai/mistral-large-2407 lyrics, verdicts
- openai/gpt-4o-2024-08-06 v deepseek/deepseek-r1 lyrics, verdicts
- google/gemini-2.0-flash-001 v microsoft/wizardlm-2-8x22b lyrics, verdicts
- google/gemini-2.5-pro v microsoft/mai-ds-r1 lyrics, verdicts
- aion-labs/aion-1.0 v qwen/qwq-32b lyrics, verdicts
- openai/gpt-5 v microsoft/phi-4-multimodal-instruct lyrics, verdicts
- openai/gpt-4.1-mini v qwen/qwen3-235b-a22b lyrics, verdicts
- google/gemini-2.5-flash-lite v deepcogito/cogito-v2-preview-llama-109b-moe lyrics, verdicts
- google/gemma-3-12b-it v qwen/qwen3-30b-a3b lyrics, verdicts
- anthropic/claude-3.7-sonnet v openai/o4-mini lyrics, verdicts
- openai/gpt-oss-20b v cognitivecomputations/dolphin3.0-r1-mistral-24b lyrics, verdicts
- openai/gpt-4o-mini v bytedance/seed-oss-36b-instruct lyrics, verdicts
- openai/gpt-oss-120b v qwen/qwen3-14b lyrics, verdicts
- meituan/longcat-flash-chat v baidu/ernie-4.5-vl-424b-a47b lyrics, verdicts
- z-ai/glm-4.5 v openai/o3-pro lyrics, verdicts
- qwen/qwen3-235b-a22b-2507 v google/gemma-3-4b-it lyrics, verdicts
- mistralai/mistral-nemo v nvidia/llama-3.1-nemotron-ultra-253b-v1 lyrics, verdicts
- openai/gpt-4-turbo v qwen/qwen3-8b lyrics, verdicts
- anthropic/claude-3.5-sonnet-20240620 v nousresearch/hermes-3-llama-3.1-70b lyrics, verdicts
- google/gemini-2.5-flash-lite-preview-06-17 v openai/gpt-3.5-turbo-16k lyrics, verdicts
- google/gemma-3-27b-it v cohere/command-r-plus-08-2024 lyrics, verdicts
- deepseek/deepseek-r1-0528 v qwen/qwen3-max lyrics, verdicts
- nousresearch/hermes-4-70b v meta-llama/llama-3.1-405b-instruct lyrics, verdicts
- deepseek/deepseek-v3.1-terminus v mistralai/mixtral-8x7b-instruct lyrics, verdicts
- qwen/qwen-plus v sao10k/l3-lunaris-8b lyrics, verdicts
- x-ai/grok-3-mini v openai/gpt-4-1106-preview lyrics, verdicts
- perplexity/sonar-reasoning-pro v perplexity/sonar-deep-research lyrics, verdicts
- anthropic/claude-3-opus v cohere/command-r7b-12-2024 lyrics, verdicts
- anthropic/claude-3.5-sonnet v liquid/lfm-7b lyrics, verdicts
- qwen/qwen-max v minimax/minimax-01 lyrics, verdicts
- deepseek/deepseek-chat v mistralai/mistral-7b-instruct lyrics, verdicts
- nvidia/llama-3.1-nemotron-70b-instruct v mistralai/mistral-medium-3.1 lyrics, verdicts
- openai/gpt-4o v stepfun-ai/step3 lyrics, verdicts
- cohere/command-r-08-2024 v arcee-ai/virtuoso-large lyrics, verdicts
- qwen/qwen3-235b-a22b-thinking-2507 v mancer/weaver lyrics, verdicts
- google/gemini-2.5-flash-lite-preview-09-2025 v openai/o4-mini-high lyrics, verdicts
- alibaba/tongyi-deepresearch-30b-a3b v mistralai/mistral-small-3.1-24b-instruct lyrics, verdicts
- tngtech/deepseek-r1t-chimera v openai/o1-pro lyrics, verdicts
- qwen/qwen-plus-2025-07-28 v openai/chatgpt-4o-latest lyrics, verdicts
- meta-llama/llama-4-scout v qwen/qwen3-30b-a3b-thinking-2507 lyrics, verdicts
- deepseek/deepseek-v3.1-base v openai/gpt-4-0314 lyrics, verdicts
- qwen/qwen3-32b v google/gemini-2.5-pro-preview lyrics, verdicts
Round 0
- anthropic/claude-sonnet-4 v nvidia/nemotron-nano-9b-v2 lyrics, verdicts
- deepseek/deepseek-chat-v3.1 v openai/gpt-4o-2024-08-06 lyrics, verdicts
- google/gemini-2.0-flash-001 v openai/gpt-3.5-turbo lyrics, verdicts
- google/gemini-2.5-pro v mistralai/mistral-large-2411 lyrics, verdicts
- deepseek/deepseek-chat-v3-0324 v aion-labs/aion-1.0 lyrics, verdicts
- openai/gpt-5 v google/gemma-2-9b-it lyrics, verdicts
- openai/gpt-4.1-mini v liquid/lfm-3b lyrics, verdicts
- google/gemini-2.5-flash-lite v deepseek/deepseek-r1-distill-qwen-32b lyrics, verdicts
- google/gemma-3-12b-it v mistralai/ministral-3b lyrics, verdicts
- anthropic/claude-3.7-sonnet v qwen/qwen-turbo lyrics, verdicts
- openai/gpt-oss-20b v meta-llama/llama-3.2-1b-instruct lyrics, verdicts
- openai/gpt-4o-mini v amazon/nova-lite-v1 lyrics, verdicts
- openai/gpt-oss-120b v baidu/ernie-4.5-300b-a47b lyrics, verdicts
- openai/gpt-5-mini v meituan/longcat-flash-chat lyrics, verdicts
- z-ai/glm-4.5 v perplexity/sonar-pro lyrics, verdicts
- qwen/qwen3-235b-a22b-2507 v meta-llama/llama-3-70b-instruct lyrics, verdicts
- mistralai/mistral-nemo v undi95/remm-slerp-l2-13b lyrics, verdicts
- openai/gpt-4.1-nano v openai/gpt-4-turbo lyrics, verdicts
- meta-llama/llama-3.3-70b-instruct v anthropic/claude-3.5-sonnet-20240620 lyrics, verdicts
- google/gemini-2.5-flash-lite-preview-06-17 v nousresearch/hermes-4-405b lyrics, verdicts
- google/gemma-3-27b-it v minimax/minimax-m1 lyrics, verdicts
- deepseek/deepseek-r1-0528 v amazon/nova-micro-v1 lyrics, verdicts
- x-ai/grok-4 v nousresearch/hermes-4-70b lyrics, verdicts
- deepseek/deepseek-v3.1-terminus v thudm/glm-4.1v-9b-thinking lyrics, verdicts
- z-ai/glm-4.5-air v qwen/qwen-plus lyrics, verdicts
- x-ai/grok-3-mini v mistralai/mistral-large lyrics, verdicts
- google/gemini-2.0-flash-lite-001 v perplexity/sonar-reasoning-pro lyrics, verdicts
- openai/gpt-5-chat v anthropic/claude-3-opus lyrics, verdicts
- anthropic/claude-3.5-sonnet v qwen/qwen-2.5-vl-7b-instruct lyrics, verdicts
- openai/gpt-5-nano v qwen/qwen-max lyrics, verdicts
- deepseek/deepseek-chat v cognitivecomputations/dolphin3.0-mistral-24b lyrics, verdicts
- openai/gpt-4o-mini-2024-07-18 v nvidia/llama-3.1-nemotron-70b-instruct lyrics, verdicts
- openai/gpt-4o v mistralai/mixtral-8x22b-instruct lyrics, verdicts
- meta-llama/llama-3.2-3b-instruct v cohere/command-r-08-2024 lyrics, verdicts
- qwen/qwen3-235b-a22b-thinking-2507 v arcee-ai/afm-4.5b lyrics, verdicts
- google/gemini-2.5-flash-lite-preview-09-2025 v neversleep/llama-3.1-lumimaid-8b lyrics, verdicts
- alibaba/tongyi-deepresearch-30b-a3b v microsoft/phi-3.5-mini-128k-instruct lyrics, verdicts
- tngtech/deepseek-r1t-chimera v mistralai/mistral-medium-3 lyrics, verdicts
- mistralai/mistral-small-24b-instruct-2501 v qwen/qwen-plus-2025-07-28 lyrics, verdicts
- meta-llama/llama-4-scout v nousresearch/deephermes-3-llama-3-8b-preview lyrics, verdicts
- qwen/qwen3-next-80b-a3b-instruct v deepseek/deepseek-v3.1-base lyrics, verdicts
- qwen/qwen3-32b v mistralai/mistral-small lyrics, verdicts
- meta-llama/llama-3.1-70b-instruct v openai/gpt-4o-2024-05-13 lyrics, verdicts
- meta-llama/llama-3.1-8b-instruct v mistralai/mistral-large-2407 lyrics, verdicts
- deepseek/deepseek-r1 v openai/gpt-4 lyrics, verdicts
- microsoft/wizardlm-2-8x22b v neversleep/noromaid-20b lyrics, verdicts
- microsoft/mai-ds-r1 v ai21/jamba-mini-1.7 lyrics, verdicts
- qwen/qwq-32b v openai/gpt-3.5-turbo-0613 lyrics, verdicts
- openai/gpt-4o-2024-11-20 v microsoft/phi-4-multimodal-instruct lyrics, verdicts
- qwen/qwen3-235b-a22b v anthracite-org/magnum-v4-72b lyrics, verdicts
- qwen/qwen-2.5-7b-instruct v deepcogito/cogito-v2-preview-llama-109b-moe lyrics, verdicts
- qwen/qwen3-30b-a3b v mistralai/mistral-7b-instruct-v0.3 lyrics, verdicts
- openai/o4-mini v mistralai/magistral-medium-2506 lyrics, verdicts
- anthropic/claude-3-haiku v cognitivecomputations/dolphin3.0-r1-mistral-24b lyrics, verdicts
- nousresearch/hermes-3-llama-3.1-405b v bytedance/seed-oss-36b-instruct lyrics, verdicts
- qwen/qwen3-14b v google/gemma-2-27b-it lyrics, verdicts
- qwen/qwen-2.5-72b-instruct v baidu/ernie-4.5-vl-424b-a47b lyrics, verdicts
- qwen/qwen3-30b-a3b-instruct-2507 v openai/o3-pro lyrics, verdicts
- google/gemma-3-4b-it v microsoft/phi-3-mini-128k-instruct lyrics, verdicts
- gryphe/mythomax-l2-13b v nvidia/llama-3.1-nemotron-ultra-253b-v1 lyrics, verdicts
- qwen/qwen3-8b v inception/mercury lyrics, verdicts
- nousresearch/hermes-3-llama-3.1-70b v ai21/jamba-large-1.7 lyrics, verdicts
- mistralai/mistral-tiny v openai/gpt-3.5-turbo-16k lyrics, verdicts
- google/gemma-3n-e4b-it v cohere/command-r-plus-08-2024 lyrics, verdicts
- qwen/qwen3-max v openai/gpt-4-turbo-preview lyrics, verdicts
- meta-llama/llama-3.1-405b-instruct v alpindale/goliath-120b lyrics, verdicts
- mistralai/mixtral-8x7b-instruct v microsoft/phi-3-medium-128k-instruct lyrics, verdicts
- sao10k/l3-lunaris-8b v baidu/ernie-4.5-21b-a3b lyrics, verdicts
- deepseek/deepseek-v3.2-exp v openai/gpt-4-1106-preview lyrics, verdicts
- perplexity/sonar-deep-research v mistralai/magistral-small-2506 lyrics, verdicts
- thedrummer/rocinante-12b v cohere/command-r7b-12-2024 lyrics, verdicts
- liquid/lfm-7b v openai/gpt-3.5-turbo-instruct lyrics, verdicts
- minimax/minimax-01 v nousresearch/deephermes-3-mistral-24b-preview lyrics, verdicts
- mistralai/mistral-7b-instruct v meta-llama/llama-3.1-405b lyrics, verdicts
- mistralai/mistral-medium-3.1 v neversleep/llama-3-lumimaid-70b lyrics, verdicts
- nousresearch/hermes-2-pro-llama-3-8b v stepfun-ai/step3 lyrics, verdicts
- thedrummer/skyfall-36b-v2 v arcee-ai/virtuoso-large lyrics, verdicts
- meta-llama/llama-3-8b-instruct v mancer/weaver lyrics, verdicts
- openai/o4-mini-high v baidu/ernie-4.5-vl-28b-a3b lyrics, verdicts
- mistralai/mistral-small-3.1-24b-instruct v mistralai/mistral-7b-instruct-v0.1 lyrics, verdicts
- anthropic/claude-3.5-haiku-20241022 v openai/o1-pro lyrics, verdicts
- openai/chatgpt-4o-latest v inflection/inflection-3-productivity lyrics, verdicts
- qwen/qwen3-30b-a3b-thinking-2507 v inflection/inflection-3-pi lyrics, verdicts
- mistralai/ministral-8b v openai/gpt-4-0314 lyrics, verdicts
- microsoft/phi-4 v google/gemini-2.5-pro-preview lyrics, verdicts
SSS: qwen/qwen-plus
SS: openai/o3-pro
A: qwen/qwen3-235b-a22b-thinking-2507, openai/gpt-4-1106-preview, qwen/qwen3-30b-a3b-thinking-2507
B: deepseek/deepseek-v3.1-terminus, mistralai/mistral-medium-3.1, openai/o4-mini, qwen/qwen-max, qwen/qwen3-max
C: openai/gpt-5, openai/chatgpt-4o-latest, deepseek/deepseek-r1, google/gemini-2.5-flash-lite-preview-06-17, openai/gpt-4o-2024-05-13, nvidia/llama-3.1-nemotron-ultra-253b-v1, bytedance/seed-oss-36b-instruct, anthropic/claude-3-opus, qwen/qwen3-32b, qwen/qwen3-14b, anthropic/claude-3.5-sonnet
D: alibaba/tongyi-deepresearch-30b-a3b, stepfun-ai/step3, tngtech/deepseek-r1t-chimera, openai/o4-mini-high, anthropic/claude-sonnet-4, microsoft/wizardlm-2-8x22b, meituan/longcat-flash-chat, qwen/qwen3-235b-a22b, arcee-ai/virtuoso-large, meta-llama/llama-3.1-405b-instruct, deepseek/deepseek-chat, anthropic/claude-3.5-sonnet-20240620, openai/gpt-4-turbo, deepseek/deepseek-v3.1-base, qwen/qwen3-30b-a3b, google/gemma-3-27b-it, microsoft/mai-ds-r1, perplexity/sonar-reasoning-pro, qwen/qwq-32b, google/gemini-2.5-flash-lite, openai/gpt-oss-20b
F: perplexity/sonar-deep-research, google/gemma-3-4b-it, nousresearch/hermes-3-llama-3.1-70b, openai/gpt-4o-mini, nvidia/llama-3.1-nemotron-70b-instruct, google/gemma-3-12b-it, z-ai/glm-4.5, cohere/command-r7b-12-2024, google/gemini-2.5-pro-preview, qwen/qwen-plus-2025-07-28, nousresearch/hermes-4-70b, mistralai/mistral-large-2407, openai/gpt-3.5-turbo-16k, anthropic/claude-3.7-sonnet, openai/gpt-4-0314, deepcogito/cogito-v2-preview-llama-109b-moe, google/gemini-2.5-flash-lite-preview-09-2025, baidu/ernie-4.5-vl-424b-a47b, google/gemini-2.0-flash-001, cognitivecomputations/dolphin3.0-r1-mistral-24b, qwen/qwen3-8b, meta-llama/llama-4-scout, sao10k/l3-lunaris-8b, openai/gpt-4o, cohere/command-r-08-2024, openai/gpt-4o-2024-08-06, minimax/minimax-01, liquid/lfm-7b, deepseek/deepseek-r1-0528, x-ai/grok-4-fast, openai/o1-pro, google/gemini-2.5-pro, openai/gpt-oss-120b, microsoft/phi-4-multimodal-instruct, aion-labs/aion-1.0, mistralai/mistral-7b-instruct, mistralai/mistral-nemo, openai/gpt-4.1 Merge branch 'master' of github.com:vadim0x60/rapbench -mini, x-ai/grok-3-mini, mistralai/mistral-small-3.1-24b-instruct, cohere/command-r-plus-08-2024, mistralai/mixtral-8x7b-instruct, mancer/weaver
FF: microsoft/phi-4, openai/gpt-3.5-turbo-0613, openai/gpt-5-chat, anthropic/claude-3-haiku, mistralai/ministral-3b, inflection/inflection-3-pi, nousresearch/hermes-2-pro-llama-3-8b, qwen/qwen-2.5-vl-7b-instruct, baidu/ernie-4.5-21b-a3b, nousresearch/hermes-3-llama-3.1-405b, baidu/ernie-4.5-300b-a47b, thedrummer/skyfall-36b-v2, qwen/qwen-2.5-72b-instruct, qwen/qwen-turbo, openai/gpt-4, inception/mercury, alpindale/goliath-120b, google/gemini-2.0-flash-lite-001, neversleep/noromaid-20b, undi95/remm-slerp-l2-13b, perplexity/sonar-pro, mistralai/magistral-small-2506, openai/gpt-5-mini, baidu/ernie-4.5-vl-28b-a3b, google/gemma-3n-e4b-it, microsoft/phi-3-medium-128k-instruct, z-ai/glm-4.5-air, openai/gpt-4o-mini-2024-07-18, openai/gpt-3.5-turbo-instruct, nvidia/nemotron-nano-9b-v2, amazon/nova-lite-v1, gryphe/mythomax-l2-13b, liquid/lfm-3b, neversleep/llama-3.1-lumimaid-8b, nousresearch/hermes-4-405b, mistralai/mistral-large, mistralai/mistral-large-2411, nousresearch/deephermes-3-llama-3-8b-preview, anthropic/claude-3.5-haiku-20241022, nousresearch/deephermes-3-mistral-24b-preview, mistralai/magistral-medium-2506, mistralai/mistral-tiny, anthracite-org/magnum-v4-72b, deepseek/deepseek-v3.2-exp, x-ai/grok-4, openai/gpt-4o-2024-11-20, microsoft/phi-3.5-mini-128k-instruct, meta-llama/llama-3.1-8b-instruct, minimax/minimax-m1, openai/gpt-4.1-nano, microsoft/phi-3-mini-128k-instruct, mistralai/mistral-7b-instruct-v0.1, thudm/glm-4.1v-9b-thinking, openai/gpt-5-nano, mistralai/mistral-small-24b-instruct-2501, meta-llama/llama-3-70b-instruct, openai/gpt-4-turbo-preview, mistralai/ministral-8b, mistralai/mistral-small, deepseek/deepseek-chat-v3.1, thedrummer/rocinante-12b, qwen/qwen3-next-80b-a3b-instruct, mistralai/mistral-medium-3, meta-llama/llama-3.2-3b-instruct, cognitivecomputations/dolphin3.0-mistral-24b, ai21/jamba-mini-1.7, google/gemma-2-9b-it, mistralai/mistral-7b-instruct-v0.3, neversleep/llama-3-lumimaid-70b, google/gemma-2-27b-it, meta-llama/llama-3-8b-instruct, amazon/nova-micro-v1, mistralai/mixtral-8x22b-instruct, arcee-ai/afm-4.5b, ai21/jamba-large-1.7, openai/gpt-3.5-turbo, meta-llama/llama-3.1-405b, deepseek/deepseek-chat-v3-0324, deepseek/deepseek-r1-distill-qwen-32b, meta-llama/llama-3.3-70b-instruct, qwen/qwen-2.5-7b-instruct, qwen/qwen3-30b-a3b-instruct-2507, meta-llama/llama-3.2-1b-instruct, inflection/inflection-3-productivity, meta-llama/llama-3.1-70b-instruct
This repository is a snakemake workflow. To reproduce the entire tournament, clone it, delete the tournament directory, set OPENROUTER_API_KEY environment variable, and run:
snakemake -c allThis will cost you around $30.