zack3d/keychecker
1
0
Fork 0
mirror of https://github.com/cunnymessiah/keychecker.git synced 2026-05-10 18:39:04 -07:00
Code Issues Packages Projects Releases Wiki Activity
Files
122eadc6bdbbeec9f31e94878f98c4692d5b5718
keychecker/OpenAI.py
T

388 lines
14 KiB
Python
Raw Blame History

import copy
import APIKey
import asyncio
oai_api_url = "https://api.openai.com/v1"
# free tier keys are no longer considered human
oai_tiers = {
'Tier 1': {'tpm': 500000, 'rpm': 500}, # why is this grayed out on the website
'Tier 2': {'tpm': 1000000, 'rpm': 5000},
'Tier 3': {'tpm': 2000000, 'rpm': 5000},
'Tier 4': {'tpm': 4000000, 'rpm': 10000},
'Tier 5': {'tpm': 40000000, 'rpm': 15000} # non gpt-5 -> 30000000, 10000
}
standard_model_ids = {
"omni-moderation-2024-09-26",
"gpt-4o-mini-audio-preview-2024-12-17",
"dall-e-3",
"dall-e-2",
"gpt-4o-audio-preview-2024-10-01",
"o1",
"gpt-4o-audio-preview",
"gpt-4o-mini-realtime-preview-2024-12-17",
"o1-2024-12-17",
"gpt-4-0314",
"gpt-4o-mini-realtime-preview",
"o1-mini-2024-09-12",
"o1-preview-2024-09-12",
"o1-mini",
"o1-preview",
"gpt-4o-mini-audio-preview",
"whisper-1",
"gpt-4-turbo",
"gpt-4o-realtime-preview-2024-10-01",
"gpt-4",
"babbage-002",
"gpt-4-turbo-preview",
"tts-1-hd-1106",
"gpt-4-0125-preview",
"gpt-4o-audio-preview-2024-12-17",
"tts-1-hd",
"gpt-4o-mini-2024-07-18",
"gpt-4o-2024-08-06",
"gpt-4o",
"tts-1",
"tts-1-1106",
"gpt-4-turbo-2024-04-09",
"davinci-002",
"gpt-3.5-turbo-1106",
"gpt-4o-mini",
"gpt-4o-2024-05-13",
"gpt-3.5-turbo-instruct",
"chatgpt-4o-latest",
"gpt-3.5-turbo-instruct-0914",
"gpt-3.5-turbo-0125",
"gpt-4o-realtime-preview-2024-12-17",
"gpt-3.5-turbo",
"gpt-3.5-turbo-16k-0613",
"gpt-4o-realtime-preview",
"gpt-3.5-turbo-16k",
"text-embedding-3-small",
"gpt-4o-2024-11-20",
"gpt-4-1106-preview",
"text-embedding-ada-002",
"text-embedding-3-large",
"o3-mini-2025-01-31",
"gpt-4-0613",
"o3-mini",
"omni-moderation-latest",
"gpt-4-base",
"o1-pro",
"o1-pro-2025-03-19",
"gpt-4o-transcribe",
"gpt-4o-transcribe-diarize",
"computer-use-preview",
"computer-use-preview-2025-03-11",
"gpt-4o-search-preview",
"gpt-4o-search-preview-2025-03-11",
"gpt-4o-mini-search-preview",
"gpt-4o-mini-search-preview-2025-03-11",
"gpt-4o-mini-transcribe",
"gpt-4o-mini-tts",
"o3",
"o4-mini",
"o3-2025-04-16",
"o4-mini-2025-04-16",
"gpt-4.1-mini",
"gpt-4.1-mini-2025-04-14",
"gpt-4.1-nano",
"gpt-4.1-nano-2025-04-14",
"gpt-4.1",
"gpt-4.1-2025-04-14",
"gpt-image-1",
"gpt-image-1-mini",
"codex-mini-latest",
"gpt-4o-realtime-preview-2025-06-03",
"gpt-4o-audio-preview-2025-06-03",
"o3-pro",
"o3-pro-2025-06-10",
"o3-deep-research",
"o3-deep-research-2025-06-26",
"o4-mini-deep-research",
"o4-mini-deep-research-2025-06-26",
"gpt-5-mini",
"gpt-5-mini-2025-08-07",
"gpt-5-nano",
"gpt-5-nano-2025-08-07",
"gpt-5",
"gpt-5-2025-08-07",
"gpt-5-chat-latest",
"gpt-5-codex",
"gpt-5-pro",
"gpt-5-pro-2025-10-06",
"gpt-5-search-api",
"gpt-5-search-api-2025-10-14",
"gpt-audio-2025-08-28",
"gpt-realtime-2025-08-28",
"gpt-audio",
"gpt-audio-mini",
"gpt-audio-mini-2025-10-06",
"gpt-realtime",
"gpt-realtime-mini",
"gpt-realtime-mini-2025-10-06",
"sora-2",
"sora-2-pro",
"gpt-5.1-2025-11-13",
"gpt-5.1-chat-latest",
"gpt-5.1",
"gpt-5.1-codex",
"gpt-5.1-codex-mini",
}
running_org_verify = False
non_slop_standard = {"gpt-5", "gpt-5-chat-latest", "o3", "gpt-4.1", "chatgpt-4o-latest", "gpt-4o"} # i imagine this is all people care about now
async def get_oai_model(key: APIKey, session, retries, org=None):
accessible_models = set()
for _ in range(retries):
headers = {'Authorization': f'Bearer {key.api_key}'}
if org is not None:
headers['OpenAI-Organization'] = org
async with session.get(f'{oai_api_url}/models', headers=headers) as response:
if response.status == 200:
data = await response.json()
for model in data["data"]:
model_id = model["id"]
if "ft:" in model_id:
key.has_special_models = True
elif model_id not in standard_model_ids and ":ft-" not in model_id:
key.extra_models = True
key.extra_model_list.add(model_id)
if model_id == "gpt-4-base" or model_id == "gpt-5-alpha-max" or model_id == "gpt-4.5-preview":
key.the_one = True
key.slop = False
if model_id == "gpt-4-32k" or model_id == "gpt-4-32k-0613":
key.real_32k = True
key.slop = False
if model_id in non_slop_standard:
key.slop = False
key.model = model_id
accessible_models.add(model_id)
key.missing_models = non_slop_standard.difference(accessible_models)
return True
elif response.status == 403:
key.model = "gpt-5"
key.access_to_model_listing = False
return True
elif response.status != 502:
return
await asyncio.sleep(0.5)
async def get_oai_key_attribs(key: APIKey, session, retries, org=None, check_id=False):
param = "max_tokens" if "gpt-4" in key.model else "max_completion_tokens"
chat_object = {"model": key.model, "messages": [{"role": "user", "content": ""}], param: 0}
for _ in range(retries):
headers = {'Authorization': f'Bearer {key.api_key}', 'accept': 'application/json'}
if org is not None:
headers['OpenAI-Organization'] = org
async with session.post(f'{oai_api_url}/chat/completions', headers=headers, json=chat_object) as response:
if (response.status == 403):
return True
elif response.status in [400, 429]:
data = await response.json()
message = data["error"]["type"]
if message is None:
return
match message:
case "access_terminated":
return
case "billing_not_active":
return
case "insufficient_quota":
key.has_quota = False
case "invalid_request_error":
key.has_quota = True
key.rpm = int(response.headers.get("x-ratelimit-limit-requests"))
tpm = int(response.headers.get("x-ratelimit-limit-tokens"))
key.tier = await get_oai_key_tier(tpm, key.model)
if check_id:
global running_org_verify
running_org_verify = True
key.id_verified = await check_id_verified(key, session, retries)
key.slop = False
return True
elif response.status not in [502, 500]:
return
await asyncio.sleep(0.5)
async def check_id_verified(key: APIKey, session, retries, org=None):
# valid parameter check is ran before unverified org detection, so the request has to be legitimate.
chat_object = {"model": "o3", "messages": [{"role": "user", "content": ""}], "max_completion_tokens": 1, "stream": True}
for attempt in range(retries):
headers = {'Authorization': f'Bearer {key.api_key}', 'accept': 'application/json'}
if org is not None:
headers['OpenAI-Organization'] = org
async with session.post(f'{oai_api_url}/chat/completions', headers=headers, json=chat_object) as response:
if response.status == 200:
return True
if response.status == 429:
continue
if attempt == retries - 1:
return False
if response.status == 400:
data = await response.json()
message = data["error"]["message"]
if "Your organization must be verified" in message:
return False
else:
return True
await asyncio.sleep(0.5)
return False
# this will weed out fake t4/t5 keys reporting a 10k rpm limit, those keys would have requested to have their rpm increased
unknown = "Tier Unknown (strange or absent token limit in header)"
async def get_oai_key_tier(tpm, model_name):
if not isinstance(tpm, int):
return unknown
fixed_tiers = copy.deepcopy(oai_tiers)
if "gpt-5" not in model_name:
fixed_tiers['Tier 5']['tpm'] = 30000000
for tier_name, tier_data in fixed_tiers.items():
if tier_data['tpm'] == tpm:
return tier_name
return unknown
async def get_oai_org(key: APIKey, session, retries):
for _ in range(retries):
async with session.get(f'{oai_api_url}/me', headers={'Authorization': f'Bearer {key.api_key}'}) as response:
if response.status == 200:
data = await response.json()
orgs = data["orgs"]["data"]
for org in orgs:
if not org["personal"]:
if org["is_default"]:
key.default_org = org["name"]
key.organizations.append(org["name"])
return True
elif response.status == 403:
key.default_org = None
return True
elif response.status != 502:
return
await asyncio.sleep(0.5)
async def clone_key(key: APIKey, session, retries):
cloned_keys = set()
if len(key.organizations) <= 0:
return False
for org in key.organizations:
if org == key.default_org:
continue
new_key = key.clone()
new_key.default_org = org
results = await asyncio.gather(get_oai_model(new_key, session, retries, org), get_oai_key_attribs(new_key, session, retries, org))
model_result, attribs_result = results
if model_result is not None and attribs_result is not None:
cloned_keys.add(new_key)
return cloned_keys
def check_manual_increase(key: APIKey):
fixed_tiers = copy.deepcopy(oai_tiers)
if "gpt-5" not in key.model:
fixed_tiers['Tier 5']['rpm'] = 10000
if key.tier in fixed_tiers:
standard_rpm = fixed_tiers[key.tier]['rpm']
return key.rpm > standard_rpm
return False
def format_key_details(key: APIKey):
details = [key.api_key]
if key.access_to_model_listing is False:
details.append("no access to /models endpoint")
if len(key.missing_models) > 0 and not key.slop and key.has_quota:
details.append(f"!!!WARNING!!! key doesn't have access to models: {key.missing_models}")
if key.default_org:
details.append(f"default org - {key.default_org}")
other_orgs = [org for org in key.organizations if org != key.default_org]
if other_orgs:
details.append(f"other orgs - {other_orgs}")
if key.has_quota:
details.append(f"{key.rpm} RPM")
if check_manual_increase(key):
details.append("(RPM increased via request)")
if key.the_one:
details.append(f"!!!god key!!!")
if key.real_32k:
details.append(f"!!!32k key!!!")
if key.has_special_models:
details.append("key has finetuned models")
if key.extra_models:
details.append(f"key has access to non-standard models: {', '.join(key.extra_model_list)}")
if key.id_verified:
details.append("id verified")
return " | ".join(details)
def pretty_print_oai_keys(keys, cloned_keys):
print('-' * 90)
group = {
"quota": {},
"no_quota": [],
"slop": []
}
for key in keys:
if key.slop:
group["slop"].append(key)
elif key.has_quota:
if key.tier not in group["quota"]:
group["quota"][key.tier] = []
group["quota"][key.tier].append(key)
else:
group["no_quota"].append(key)
quota_keys_by_tier = group["quota"]
stats = {
"quota_count": sum(len(tier_keys) for tier_keys in quota_keys_by_tier.values()),
"no_quota_count": len(group["no_quota"]),
"slop_count": len(group["slop"]),
"org_count": sum(1 for k in keys if k.organizations),
"verified_org_count": sum(1 for k in keys if k.id_verified),
"t5_count": len(quota_keys_by_tier.get('Tier 5', []))
}
if stats["quota_count"] > 0:
print(f"\n--- Verified {stats['quota_count']} OpenAI keys with quota ---")
for tier_name in reversed(oai_tiers.keys()):
if tier_name in quota_keys_by_tier:
print(f"\nFound {len(quota_keys_by_tier[tier_name])} keys in {tier_name}:")
for key in quota_keys_by_tier[tier_name]:
print(format_key_details(key))
if unknown in quota_keys_by_tier:
print(f"\nFound {len(quota_keys_by_tier[unknown])} keys of {unknown}:")
for key in quota_keys_by_tier[unknown]:
print(format_key_details(key))
if stats["no_quota_count"] > 0:
print(f"\n--- Verified {stats['no_quota_count']} OpenAI keys with no quota ---")
for key in group["no_quota"]:
print(format_key_details(key))
if stats["slop_count"] > 0:
print(f"\n--- Verified {stats['slop_count']} OpenAI keys that are slop (no access to useful models or on free tier) ---")
for key in group["slop"]:
print(format_key_details(key))
if cloned_keys:
print(f'\n--- Cloned {len(cloned_keys)} keys due to finding alternative orgs that could prompt ---')
summary = f'\n--- Total Valid OpenAI Keys: {len(keys)} ' + f'({stats["quota_count"]} in quota, {stats["no_quota_count"]} no quota, {stats["slop_count"]} slop, {stats["org_count"]} orgs'
if running_org_verify:
summary += f' - {stats["verified_org_count"]} with id verified'
summary += f', {stats["t5_count"]} Tier 5) ---\n'
print(summary)
Reference in New Issue View Git Blame Copy Permalink