Model Request Ratios (Legacy)

Legacy plans only: This page applies to existing subscribers on retired request-count plans (Promo, Lite, Pro, Pro+, Max, and Max+), which remain active until the end of their current billing period. New subscriptions use the Minion ($20/month) and **Boss** ($80/month) plans — see Model Costs & Usage Estimates, Minion, and Boss.

Legacy plan limits at a glance

	Promo	Lite	Pro	Pro+	Max	Max+
Price (legacy)	$1/mo	$15/mo	$30/mo	$45/mo	$60/mo	$120/mo
Requests per 5-hour window	30	400	800	1,200	1,600	3,200
Requests per week	420	7,500	13,000	19,500	26,000	52,000
Burst limit (requests/minute)	10	15	25	25	25	25
Autonomous agents	—	✓	✓	✓	✓	✓

Not every model costs the same slice of a legacy request-count plan. Each model has a ratio — how much of your allowance one call to it uses. Auto and flash models are the 1.0 floor. GLM-5.2 costs 2.0 — one call uses two requests’ worth. Your plan’s “requests per window” number is what that buys on Auto and other 1.0 models; GLM-5.2 calls consume twice as much. On Promo (30 per 5-hour window) that is 30 Auto calls or 15 GLM-5.2 calls; a very long DeepSeek-V4.1-Flash or GLM-5.3-Flash prompt costs 3.0 (down to 10 on Promo).

Note: mid-tier models sit between 1.0 and 2.0; premium chat models cost more — see the table below. Long-context surcharges on included models are separate — see Long-context surcharge. Image models use a fixed ratio per image — see Image generation ratios.

Promo plan ($1/mo, retired; existing subscribers only): Includes a 2-day trial on all models. After trial, Qwen3.8 27b and Llama 3.2 3B receive a special 0.1 request ratio (delivering 300 requests every 5 hours), and GPT-OSS-120B costs 0.5 (60 requests every 5 hours).

Ratios at a glance

Ratio	Models
20	`GLM-5.3` when the prompt exceeds 262,144 tokens
3	`DeepSeek-V4.1-Flash`, `DeepSeek-V4.1-Flash-1017`, `GLM-5.3-Flash` when the prompt exceeds 262,144 tokens
2.5	`DeepSeek-V4-Flash-0731` when the prompt exceeds 262,144 tokens
10	`GLM-5.3`
2	`DeepSeek-V4.1-Flash`, `DeepSeek-V4.1-Flash-1017`, `GLM-5.3-Flash`
1.7	`Agnes-3.0-Flash`, `DeepSeek-V4-Flash-0731`
1	`Auto`
0.1	`Gemma 4 26B`, `GLM-4.7-Flash`, `GPT-OSS-120B`, `Llama 3.2 3B`, `MiniMax M3`, `Nemotron-3-120B`, `Qwen3.8 27b`

Peak hours (06:00–10:00 UTC): GLM-5.3 costs more of your allowance — see Peak-hour ratios.

Use the model name exactly as shown as the model value in your request. Tables below list models on subscription plans (Promo through Max+). On-demand-only models (for example GLM-5.2) are billed per token and are not shown here. Models included only on Minion and Boss (for example DeepSeek-V4-Pro-0813, Kimi K3) are billed per token on these plans and are not shown here either.

Legacy Auto Agent plans: Existing subscribers on Auto Agent Mini, Lite, Pro, or Max keep their grandfathered allowance at the same ratios until they choose to switch plans.

How many requests you get

Per 5-hour window

Model	Ratio	Promo	Lite	Pro	Pro+	Max	Max+
GLM-5.3 (prompt > 262,144)	20	1	20	40	60	80	160
GLM-5.3	10	3	40	80	120	160	320
DeepSeek-V4.1-Flash (prompt > 262,144)	3	10	133	266	400	533	1,066
DeepSeek-V4.1-Flash-1017 (prompt > 262,144)	3	10	133	266	400	533	1,066
GLM-5.3-Flash (prompt > 262,144)	3	10	133	266	400	533	1,066
DeepSeek-V4-Flash-0731 (prompt > 262,144)	2.5	12	160	320	480	640	1,280
DeepSeek-V4.1-Flash	2	15	200	400	600	800	1,600
DeepSeek-V4.1-Flash-1017	2	15	200	400	600	800	1,600
GLM-5.3-Flash	2	15	200	400	600	800	1,600
Agnes-3.0-Flash	1.7	17	235	470	705	941	1,882
DeepSeek-V4-Flash-0731	1.7	17	235	470	705	941	1,882
Auto	1	30	400	800	1,200	1,600	3,200
Gemma 4 26B	0.1	300	4,000	8,000	12,000	16,000	32,000
GLM-4.7-Flash	0.1	300	4,000	8,000	12,000	16,000	32,000
GPT-OSS-120B	0.1	300	4,000	8,000	12,000	16,000	32,000
Llama 3.2 3B	0.1	300	4,000	8,000	12,000	16,000	32,000
MiniMax M3	0.1	300	4,000	8,000	12,000	16,000	32,000
Nemotron-3-120B	0.1	300	4,000	8,000	12,000	16,000	32,000
Qwen3.8 27b	0.1	300	4,000	8,000	12,000	16,000	32,000

Per week

Weekly caps reset Monday 00:00 UTC.

Model	Ratio	Promo	Lite	Pro	Pro+	Max	Max+
GLM-5.3 (prompt > 262,144)	20	21	375	650	975	1,300	2,600
GLM-5.3	10	42	750	1,300	1,950	2,600	5,200
DeepSeek-V4.1-Flash (prompt > 262,144)	3	140	2,500	4,333	6,500	8,666	17,333
DeepSeek-V4.1-Flash-1017 (prompt > 262,144)	3	140	2,500	4,333	6,500	8,666	17,333
GLM-5.3-Flash (prompt > 262,144)	3	140	2,500	4,333	6,500	8,666	17,333
DeepSeek-V4-Flash-0731 (prompt > 262,144)	2.5	168	3,000	5,200	7,800	10,400	20,800
DeepSeek-V4.1-Flash	2	210	3,750	6,500	9,750	13,000	26,000
DeepSeek-V4.1-Flash-1017	2	210	3,750	6,500	9,750	13,000	26,000
GLM-5.3-Flash	2	210	3,750	6,500	9,750	13,000	26,000
Agnes-3.0-Flash	1.7	247	4,411	7,647	11,470	15,294	30,588
DeepSeek-V4-Flash-0731	1.7	247	4,411	7,647	11,470	15,294	30,588
Auto	1	420	7,500	13,000	19,500	26,000	52,000
Gemma 4 26B	0.1	4,200	75,000	130,000	195,000	260,000	520,000
GLM-4.7-Flash	0.1	4,200	75,000	130,000	195,000	260,000	520,000
GPT-OSS-120B	0.1	4,200	75,000	130,000	195,000	260,000	520,000
Llama 3.2 3B	0.1	4,200	75,000	130,000	195,000	260,000	520,000
MiniMax M3	0.1	4,200	75,000	130,000	195,000	260,000	520,000
Nemotron-3-120B	0.1	4,200	75,000	130,000	195,000	260,000	520,000
Qwen3.8 27b	0.1	4,200	75,000	130,000	195,000	260,000	520,000

Peak-hour ratios

Some models cost more of a legacy request-count allowance during peak hours: every day from 06:00–10:00 UTC (14:00–18:00 UTC+8). Outside that window the standard ratios apply.

Model	Standard ratio	Peak ratio
GLM-5.3 (prompt > 262,144)	20	40
GLM-5.3	10	20

The ratio is set when a request is received: a call received before the end of the window is charged the peak ratio even if it finishes later. It applies to your window and weekly allowances alike. Plans metered by tokens (Minion and Boss) have no ratios and are not affected.

Per 5-hour window at peak ratios

Model	Ratio	Promo	Lite	Pro	Pro+	Max	Max+
GLM-5.3 (prompt > 262,144)	40	0	10	20	30	40	80
GLM-5.3	20	1	20	40	60	80	160

Per week at peak ratios

Model	Ratio	Promo	Lite	Pro	Pro+	Max	Max+
GLM-5.3 (prompt > 262,144)	40	10	187	325	487	650	1,300
GLM-5.3	20	21	375	650	975	1,300	2,600

Image generation ratios

Each successful image counts as one call. Ratios apply per image, not per token.

Model	Ratio
Agnes-Image-2.5-Flash	25

You do not have to spend a window on one model — costs simply add up. On Pro, 600 GLM-5.2 calls (1,200) plus 400 Auto calls (400) is 1,600 requests’ worth.

Long-context surcharge

GLM-5.3 requests whose prompt exceeds 262,144 tokens cost 20.0 instead of 10.0. DeepSeek-V4.1-Flash and GLM-5.3-Flash requests whose prompt exceeds 262,144 tokens cost 3.0 instead of 2.0. DeepSeek-V4-Flash-0731 uses 2.5 above the same threshold instead of 1.7. During peak hours (06:00–10:00 UTC) the ratios scale as described in Peak-hour ratios: GLM-5.3 costs 20.0 (40.0 above 262,144 tokens).

The surcharge is measured on input only. A long reply to a short prompt still costs the base ratio.
The threshold is measured on upstream-reported input tokens after the call completes. X-Quota-Cost on the response uses a pre-flight estimate and may read lower until settlement; the dashboard is always exact.

If you routinely work above 262k tokens, MiniMax M3 has a larger context window and carries no long-context surcharge on plan quota.

Reading the cost at runtime

You never have to hardcode this table. GET /v1/models reports the ratio on every model:

curl https://api.openference.com/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

{
  "id": "DeepSeek-V4.1-Flash",
  "quota_multiplier": 2,
  "quota_context_surcharge": {
    "threshold_input_tokens": 262144,
    "factor": 2
  }
}

quota_context_surcharge is present only on models whose cost scales with prompt size. Models that cost more during peak hours also report quota_multiplier_peak and quota_peak_window_utc:

{
  "id": "GLM-5.3",
  "quota_multiplier": 10,
  "quota_multiplier_peak": 20,
  "quota_peak_window_utc": {
    "start_hour": 6,
    "end_hour": 10
  },
  "quota_context_surcharge": {
    "threshold_input_tokens": 262144,
    "factor": 2
  }
}

quota_multiplier is the standard ratio. quota_multiplier_peak applies every day from start_hour (inclusive) to end_hour (exclusive), UTC, on request-count plans. Both fields are present only on models with peak pricing. Every response also carries what that specific call used:

Header	Meaning
`X-Quota-Cost`	What this request cost, in requests (e.g. `1`, `2`, `2.5`), including any peak-hour ratio
`X-Quota-Remaining`	Approximately what is left in the current window

Note: X-Quota-Remaining is read when the request is admitted, so with several requests in flight at once it can read slightly high. The dashboard under Billing is always exact.

What still counts as one request

The ratio changes how much a call costs, not which calls count:

success and client_error (4xx) count.
Upstream provider errors and capacity-unavailable situations do not count — a failed request is refunded in full, at whatever it would have cost.
The per-minute burst limit is unaffected. It is a hard cap on calls per minute and is never extended by credits.

Model Costs & Usage Estimates — current token rates and request estimates for Minion and Boss
Plans & Usage — allowances, windows, and burst limits
Pricing & Transparency — on-demand token rates beyond your allowance
Model Catalog — context windows and capabilities

Getting Started

API Reference

Client Integrations

Models

Billing, Quotas & Usage

API Keys & Security

Reliability & Architecture

Troubleshooting

Reference

FAQ

Model Request Ratios (Legacy)

Model Request Ratios (Legacy)

Legacy plan limits at a glance

Ratios at a glance

How many requests you get

Per 5-hour window

Per week

Peak-hour ratios

Per 5-hour window at peak ratios

Per week at peak ratios

Image generation ratios

Long-context surcharge

Reading the cost at runtime

What still counts as one request

​Model Request Ratios (Legacy)

​Legacy plan limits at a glance

​Ratios at a glance

​How many requests you get

​Per 5-hour window

​Per week

​Peak-hour ratios

​Per 5-hour window at peak ratios

​Per week at peak ratios

​Image generation ratios

​Long-context surcharge

​Reading the cost at runtime

​What still counts as one request

​Related

Model Request Ratios (Legacy)

Legacy plan limits at a glance

Ratios at a glance

How many requests you get

Per 5-hour window

Per week

Peak-hour ratios

Per 5-hour window at peak ratios

Per week at peak ratios

Image generation ratios

Long-context surcharge

Reading the cost at runtime

What still counts as one request

Related