Falcon-7B Instruct

By Technology Innovation Institute (TII), in the Falcon family.

Released: 25 Apr 2023
Status: Legacy
Context: 2K
Max output: —
Input per 1M tokens: —
Output per 1M tokens: —
Open weights: Yes
License: apache-2.0
Parameters: 7.22B
Knowledge cutoff: —
Reasoning: No
Modalities: text → text

Category scores

Out of 100, from independent and third-party results only, by method category-2026-09-29.1.

Not enough independent results yet

Only independent and third-party results rank a model; self-reported and unattributed numbers never do.

Benchmark results

Every result we hold for this model, with who ran it and where it was published.

Knowledge & reasoning

Knowledge & reasoning results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
AI2 Reasoning Challenge (ARC)	Challenge set accuracy	42.4%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
AI2 Reasoning Challenge (ARC)	Challenge set accuracy	47.9%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
AI2 Reasoning Challenge (ARC)	Challenge set accuracy	44.5%	Self-reported	—	—	The Falcon Series of Open Language Models via Epoch AI
AI2 Reasoning Challenge (ARC)	Challenge set accuracy	43.4%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
AI2 Reasoning Challenge (ARC)	Challenge set accuracy	43.4%	Self-reported	—	—	XGen-7B Technical Report via Epoch AI
AI2 Reasoning Challenge (ARC)	Challenge set accuracy	42.4%	Self-reported	—	—	Qwen Technical Report via Epoch AI
AI2 Reasoning Challenge (ARC)	Challenge set accuracy	36.3%	Self-reported	—	—	Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI
BIG-Bench Hard (BBH)	Average accuracy	28.8%	Third party	—	—	Baichuan 2: Open Large-scale Language Models via Epoch AI
BIG-Bench Hard (BBH)	Average accuracy	28.0%	Self-reported	—	—	Qwen Technical Report via Epoch AI
HellaSwag	Accuracy	78.1%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
HellaSwag	Accuracy	78.1%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
HellaSwag	Accuracy	76.4%	Self-reported	—	—	XGen-7B Technical Report via Epoch AI
HellaSwag	Accuracy	76.3%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
HellaSwag	Accuracy	76.3%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
HellaSwag	Accuracy	74.1%	Self-reported	—	—	Qwen Technical Report via Epoch AI
HellaSwag	Accuracy	54.2%	Self-reported	—	—	Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI
LAMBADA	Accuracy	74.9%	Self-reported	—	—	The Falcon Series of Open Language Models via Epoch AI
MMLU	Accuracy	26.2%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
MMLU	Accuracy	26.0%	Third party	—	—	Baichuan 2: Open Large-scale Language Models via Epoch AI
MMLU	Accuracy	35.0%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
MMLU	Accuracy	26.9%	Self-reported	—	—	XGen-7B Technical Report via Epoch AI
MMLU	Accuracy	26.9%	Self-reported	—	—	Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI
MMLU	Accuracy	26.2%	Self-reported	—	—	Qwen Technical Report via Epoch AI
MMLU	Accuracy	23.9%	Self-reported	—	—	XGen-7B Technical Report via Epoch AI
OpenBookQA	Accuracy	51.6%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
OpenBookQA	Accuracy	44.0%	Self-reported	—	—	XGen-7B Technical Report via Epoch AI
OpenBookQA	Accuracy	32.0%	Self-reported	—	—	Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI
PIQA	Accuracy	76.7%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
PIQA	Accuracy	80.3%	Self-reported	—	—	The Falcon Series of Open Language Models via Epoch AI
PIQA	Accuracy	79.4%	Self-reported	—	—	XGen-7B Technical Report via Epoch AI
PIQA	Accuracy	79.4%	Self-reported	—	—	Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI
PIQA	Accuracy	76.7%	Self-reported	—	—	Qwen Technical Report via Epoch AI
WinoGrande	Accuracy	66.3%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
WinoGrande	Accuracy	67.2%	Self-reported	—	—	XGen-7B Technical Report via Epoch AI
WinoGrande	Accuracy	67.2%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
WinoGrande	Accuracy	67.2%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI
WinoGrande	Accuracy	66.2%	Self-reported	—	—	Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI

Mathematics

Mathematics results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
GSM8K	Exact match	6.8%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
GSM8K	Exact match	5.5%	Third party	—	—	Baichuan 2: Open Large-scale Language Models via Epoch AI
GSM8K	Exact match	4.0%	Third party	—	—	Stanford HELM via Epoch AI
GSM8K	Exact match	6.8%	Self-reported	—	—	Textbooks Are All You Need II: phi-1.5 technical report via Epoch AI
GSM8K	Exact match	6.8%	Self-reported	—	—	Qwen Technical Report via Epoch AI
GSM8K	Exact match	4.6%	Self-reported	—	—	Falcon2-11B Technical Report via Epoch AI

Factuality & hallucination

Factuality & hallucination results
Benchmark	Metric	Value	Provenance	Setting	Evaluated	Source
TriviaQA	Exact match	64.6%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
TriviaQA	Exact match	61.1%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
TriviaQA	Exact match	56.8%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI
TriviaQA	Exact match	52.6%	Third party	—	—	Llama 2: Open Foundation and Fine-Tuned Chat Models via Epoch AI

Epoch AI, 'Capabilities & benchmarking', epoch.ai/benchmarks, CC BY 4.0.

Other Falcon versions

Sources

https://huggingface.co/tiiuae/falcon-7b-instruct