@pydantic/genai-prices 0.0.67 → 0.0.68

package/dist/cli.js

@@ -105,7 +105,7 @@ class Ve {
       number: void 0,
       __: void 0,
       key: void 0
-    }, u), r = Ee(o), _ = typeof o == "string", h = De(Object.assign(/* @__PURE__ */ Object.create(null), s.alias)), k = Object.assign({
+    }, u), r = Ee(o), _ = typeof o == "string", h = Fe(Object.assign(/* @__PURE__ */ Object.create(null), s.alias)), k = Object.assign({
       "boolean-negation": !0,
       "camel-case-expansion": !0,
       "combine-arrays": !1,
@@ -171,15 +171,15 @@ class Ve {
     });
     let L = null;
     Ae();
-    let F = [];
+    let D = [];
     const b = Object.assign(/* @__PURE__ */ Object.create(null), { _: [] }), se = {};
     for (let e = 0; e < r.length; e++) {
       const a = r[e], c = a.replace(/^-{3,}/, "---");
       let n, t, l, p, d, v;
       if (a !== "--" && /^-/.test(a) && $(a))
-        Y(a);
+        Z(a);
       else if (c.match(/^---+(=|$)/)) {
-        Y(a);
+        Z(a);
         continue;
       } else if (a.match(/^--.+=/) || !k["short-option-groups"] && a.match(/^-.+=/))
         p = a.match(/^--?([^=]+)=([\s\S]*)$/), p !== null && Array.isArray(p) && p.length >= 3 && (g(p[1], m.arrays) ? e = K(e, p[1], r, p[2]) : g(p[1], m.nargs) !== !1 ? e = j(e, p[1], r, p[2]) : f(p[1], p[2], !0));
@@ -216,24 +216,24 @@ class Ve {
       } else if (a.match(/^-[0-9]$/) && a.match(y) && g(a.slice(1), m.bools))
         t = a.slice(1), f(t, N(t));
       else if (a === "--") {
-        F = r.slice(e + 1);
+        D = r.slice(e + 1);
         break;
       } else if (k["halt-at-non-option"]) {
-        F = r.slice(e);
+        D = r.slice(e);
         break;
       } else
-        Y(a);
+        Z(a);
     }
     me(b, !0), me(b, !1), we(b), fe(), ue(b, m.aliases, q, !0), be(b), k["set-placeholder-key"] && ve(b), Object.keys(m.counts).forEach(function(e) {
       O(b, e.split(".")) || f(e, 0);
-    }), G && F.length && (b[M] = []), F.forEach(function(e) {
+    }), G && D.length && (b[M] = []), D.forEach(function(e) {
       b[M].push(e);
     }), k["camel-case-expansion"] && k["strip-dashed"] && Object.keys(b).filter((e) => e !== "--" && e.includes("-")).forEach((e) => {
       delete b[e];
     }), k["strip-aliased"] && [].concat(...Object.keys(h).map((e) => h[e])).forEach((e) => {
       k["camel-case-expansion"] && e.includes("-") && delete b[e.split(".").map((a) => U(a)).join(".")], delete b[e];
     });
-    function Y(e) {
+    function Z(e) {
       const a = H("_", e);
       (typeof a == "string" || typeof a == "number") && b._.push(a);
     }
@@ -265,9 +265,9 @@ class Ve {
           t = Array.isArray(d) ? d : [d];
         }
       } else {
-        S(n) || t.push(Z(a, n, !0));
+        S(n) || t.push(Y(a, n, !0));
         for (let d = e + 1; d < c.length && !(!k["greedy-arrays"] && t.length > 0 || p && typeof p == "number" && t.length >= p || (l = c[d], /^-/.test(l) && !y.test(l) && !$(l))); d++)
-          e = d, t.push(Z(a, l, _));
+          e = d, t.push(Y(a, l, _));
       }
       return typeof p == "number" && (p && t.length < p || isNaN(p) && t.length === 0) && (L = Error(I("Not enough arguments following: %s", a))), f(a, t), e;
     }
@@ -278,7 +278,7 @@ class Ve {
         }).join(".");
         re(e, l);
       }
-      const n = Z(e, a, c), t = e.split(".");
+      const n = Y(e, a, c), t = e.split(".");
       E(b, t, n), m.aliases[e] && m.aliases[e].forEach(function(l) {
         const p = l.split(".");
         E(b, p, n);
@@ -301,8 +301,8 @@ class Ve {
     function re(e, a) {
       m.aliases[e] && m.aliases[e].length || (m.aliases[e] = [a], T[a] = !0), m.aliases[a] && m.aliases[a].length || re(a, e);
     }
-    function Z(e, a, c) {
-      c && (a = Fe(a)), (g(e, m.bools) || g(e, m.counts)) && typeof a == "string" && (a = a === "true");
+    function Y(e, a, c) {
+      c && (a = De(a)), (g(e, m.bools) || g(e, m.counts)) && typeof a == "string" && (a = a === "true");
       let n = Array.isArray(a) ? a.map(function(t) {
         return H(e, t);
       }) : H(e, a);
@@ -494,7 +494,7 @@ class Ve {
     };
   }
 }
-function De(i) {
+function Fe(i) {
   const o = [], u = /* @__PURE__ */ Object.create(null);
   let s = !0;
   for (Object.keys(i).forEach(function(r) {
@@ -524,7 +524,7 @@ function X(i) {
 function pe(i) {
   return i === "__proto__" ? "___proto___" : i;
 }
-function Fe(i) {
+function De(i) {
   return typeof i == "string" && (i[0] === "'" || i[0] === '"') && i[i.length - 1] === i[0] ? i.substring(1, i.length - 1) : i;
 }
 /**
@@ -1305,6 +1305,51 @@ const ne = [
           }
         ]
       },
+      {
+        id: "claude-sonnet-5",
+        name: "Claude Sonnet 5",
+        description: "Our most agentic Sonnet model, approaching Opus 4.8 capability at lower cost",
+        match: {
+          or: [
+            {
+              starts_with: "claude-sonnet-5"
+            },
+            {
+              starts_with: "claude-sonnet-5.0"
+            },
+            {
+              starts_with: "claude-5-sonnet"
+            },
+            {
+              starts_with: "claude-5.0-sonnet"
+            }
+          ]
+        },
+        context_window: 1e6,
+        price_comments: "Flat pricing across full 1M context window (no tiered pricing). Introductory pricing ($2/$10 per MTok) applies through 2026-08-31; standard pricing ($3/$15) applies from 2026-09-01. Ref: https://www.anthropic.com/news/claude-sonnet-5",
+        prices: [
+          {
+            prices: {
+              input_mtok: 2,
+              cache_write_mtok: 2.5,
+              cache_read_mtok: 0.2,
+              output_mtok: 10
+            }
+          },
+          {
+            constraint: {
+              start_date: "2026-09-01",
+              type: "start_date"
+            },
+            prices: {
+              input_mtok: 3,
+              cache_write_mtok: 3.75,
+              cache_read_mtok: 0.3,
+              output_mtok: 15
+            }
+          }
+        ]
+      },
       {
         id: "claude-v1",
         description: "Retired, here to match price sources",
@@ -1742,6 +1787,35 @@ const ne = [
           }
         }
       },
+      {
+        id: "global.anthropic.claude-sonnet-5-v1:0",
+        match: {
+          contains: "global.anthropic.claude-sonnet-5"
+        },
+        price_comments: "Flat pricing across full 1M context window (no tiered pricing). Promotional launch pricing ($2/$10 per MTok) through 2026-08-31; standard ($3/$15) from 2026-09-01. Ref: https://aws.amazon.com/bedrock/pricing/",
+        prices: [
+          {
+            prices: {
+              input_mtok: 2,
+              cache_write_mtok: 2.5,
+              cache_read_mtok: 0.2,
+              output_mtok: 10
+            }
+          },
+          {
+            constraint: {
+              start_date: "2026-09-01",
+              type: "start_date"
+            },
+            prices: {
+              input_mtok: 3,
+              cache_write_mtok: 3.75,
+              cache_read_mtok: 0.3,
+              output_mtok: 15
+            }
+          }
+        ]
+      },
       {
         id: "google.gemma-3-12b-it",
         name: "Gemma 3 12B IT",
@@ -2072,6 +2146,30 @@ const ne = [
           output_mtok: 0.65
         }
       },
+      {
+        id: "openai.gpt-5.4",
+        name: "GPT-5.4",
+        match: {
+          equals: "openai.gpt-5.4"
+        },
+        prices: {
+          input_mtok: 2.75,
+          cache_read_mtok: 0.275,
+          output_mtok: 16.5
+        }
+      },
+      {
+        id: "openai.gpt-5.5",
+        name: "GPT-5.5",
+        match: {
+          equals: "openai.gpt-5.5"
+        },
+        prices: {
+          input_mtok: 5.5,
+          cache_read_mtok: 0.55,
+          output_mtok: 33
+        }
+      },
       {
         id: "openai.gpt-oss-120b-1:0",
         name: "gpt-oss-120b",
@@ -2699,6 +2797,60 @@ const ne = [
             ]
           }
         }
+      },
+      {
+        id: "regional.anthropic.claude-sonnet-5-v1:0",
+        match: {
+          or: [
+            {
+              starts_with: "anthropic.claude-sonnet-5"
+            },
+            {
+              starts_with: "claude-sonnet-5"
+            },
+            {
+              contains: "us.anthropic.claude-sonnet-5"
+            },
+            {
+              contains: "au.anthropic.claude-sonnet-5"
+            },
+            {
+              contains: "apac.anthropic.claude-sonnet-5"
+            },
+            {
+              contains: "eu.anthropic.claude-sonnet-5"
+            },
+            {
+              contains: "us-gov.anthropic.claude-sonnet-5"
+            },
+            {
+              contains: "jp.anthropic.claude-sonnet-5"
+            }
+          ]
+        },
+        price_comments: "Regional/cross-region endpoints carry a 10% premium over global (AWS published only the global promo rate; regional computed as global +10%, per the documented regional premium). Promotional launch pricing through 2026-08-31; standard from 2026-09-01. Ref: https://aws.amazon.com/bedrock/pricing/",
+        prices: [
+          {
+            prices: {
+              input_mtok: 2.2,
+              cache_write_mtok: 2.75,
+              cache_read_mtok: 0.22,
+              output_mtok: 11
+            }
+          },
+          {
+            constraint: {
+              start_date: "2026-09-01",
+              type: "start_date"
+            },
+            prices: {
+              input_mtok: 3.3,
+              cache_write_mtok: 4.125,
+              cache_read_mtok: 0.33,
+              output_mtok: 16.5
+            }
+          }
+        ]
       }
     ]
   },
@@ -3991,6 +4143,20 @@ Compared to other leading proprietary and open-weights models Command A delivers
           output_mtok: 4.4
         }
       },
+      {
+        id: "glm-5p2",
+        name: "GLM-5.2",
+        description: "GLM-5.2 introduces a robust 1M-token context and advanced, multi-effort coding capabilities to significantly enhance performance on long-horizon tasks. Features a new IndexShare architecture and improved MTP layer for greater efficiency. 743B parameter MoE model from Z.ai.",
+        match: {
+          equals: "accounts/fireworks/models/glm-5p2"
+        },
+        context_window: 104e4,
+        prices: {
+          input_mtok: 1.4,
+          cache_read_mtok: 0.14,
+          output_mtok: 4.4
+        }
+      },
       {
         id: "gpt-oss-120b",
         name: "OpenAI gpt-oss-120b",
@@ -4045,6 +4211,20 @@ Compared to other leading proprietary and open-weights models Command A delivers
           output_mtok: 4
         }
       },
+      {
+        id: "kimi-k2p7-code",
+        name: "Kimi K2.7 Code",
+        description: "Kimi K2.7 Code is a coding-focused agentic model built upon Kimi K2.6, delivering substantial improvements on real-world long-horizon coding tasks while reducing thinking tokens by roughly 30% compared to its predecessor.",
+        match: {
+          equals: "accounts/fireworks/models/kimi-k2p7-code"
+        },
+        context_window: 262144,
+        prices: {
+          input_mtok: 0.95,
+          cache_read_mtok: 0.19,
+          output_mtok: 4
+        }
+      },
       {
         id: "llama-v3p1-8b-instruct",
         name: "Llama 3.1 8B Instruct",
@@ -4111,6 +4291,20 @@ Compared to other leading proprietary and open-weights models Command A delivers
           output_mtok: 1.2
         }
       },
+      {
+        id: "nemotron-3-ultra-nvfp4",
+        name: "NVIDIA Nemotron 3 Ultra NVFP4",
+        description: "Frontier-scale LLM from NVIDIA using a hybrid Latent Mixture-of-Experts (LatentMoE) architecture with interleaved Mamba-2 and MoE layers plus select Attention layers. Features 55B active parameters out of 550B total and Multi-Token Prediction layers for faster generation, optimized for complex multi-step agents, long-context analysis, and high-accuracy reasoning over code, math, and science.",
+        match: {
+          equals: "accounts/fireworks/models/nemotron-3-ultra-nvfp4"
+        },
+        context_window: 262e3,
+        prices: {
+          input_mtok: 0.6,
+          cache_read_mtok: 0.12,
+          output_mtok: 2.4
+        }
+      },
       {
         id: "qwen2p5-vl-72b-instruct",
         name: "Qwen2.5-VL 72B Instruct",
@@ -4149,6 +4343,20 @@ Compared to other leading proprietary and open-weights models Command A delivers
           cache_read_mtok: 0.1,
           output_mtok: 3
         }
+      },
+      {
+        id: "qwen3p7-plus",
+        name: "Qwen3.7 Plus",
+        description: "Qwen3.7 Plus is Alibaba's latest flagship closed model, available exclusively through Fireworks AI outside of Alibaba's own infrastructure.",
+        match: {
+          equals: "accounts/fireworks/models/qwen3p7-plus"
+        },
+        context_window: 262144,
+        prices: {
+          input_mtok: 0.4,
+          cache_read_mtok: 0.08,
+          output_mtok: 1.6
+        }
       }
     ]
   },
@@ -9560,6 +9768,65 @@ Compared to other leading proprietary and open-weights models Command A delivers
       ]
     },
     extractors: [
+      {
+        api_flavor: "default",
+        root: "usage",
+        model_path: "model",
+        mappings: [
+          {
+            path: "input_tokens",
+            dest: "input_tokens",
+            required: !0
+          },
+          {
+            path: "cache_creation_input_tokens",
+            dest: "input_tokens",
+            required: !1
+          },
+          {
+            path: "cache_read_input_tokens",
+            dest: "input_tokens",
+            required: !1
+          },
+          {
+            path: "cache_creation_input_tokens",
+            dest: "cache_write_tokens",
+            required: !1
+          },
+          {
+            path: "cache_read_input_tokens",
+            dest: "cache_read_tokens",
+            required: !1
+          },
+          {
+            path: "output_tokens",
+            dest: "output_tokens",
+            required: !0
+          }
+        ]
+      },
+      {
+        api_flavor: "responses",
+        root: "usage",
+        model_path: "model",
+        mappings: [
+          {
+            path: "input_tokens",
+            dest: "input_tokens",
+            required: !0
+          },
+          {
+            path: ["input_tokens_details", "cached_tokens"],
+            dest: "cache_read_tokens",
+            required: !1
+          },
+          {
+            path: "output_tokens",
+            dest: "output_tokens",
+            required: !0
+          }
+        ]
+      },
       {
         api_flavor: "chat",
         root: "usage",
@@ -13256,6 +13523,43 @@ Compared to other leading proprietary and open-weights models Command A delivers
           output_mtok: 15
         }
       },
+      {
+        id: "anthropic/claude-sonnet-5",
+        match: {
+          or: [
+            {
+              equals: "anthropic/claude-sonnet-5"
+            },
+            {
+              equals: "anthropic/claude-sonnet-5:beta"
+            }
+          ]
+        },
+        context_window: 1e6,
+        price_comments: "Flat pricing across full 1M context window (no tiered pricing). Introductory pricing ($2/$10 per MTok) applies through 2026-08-31; standard ($3/$15) from 2026-09-01. OpenRouter mirrors Anthropic first-party pricing; $2/$10 verified live via the OpenRouter API on 2026-06-30. Ref: https://openrouter.ai/anthropic/claude-sonnet-5",
+        prices: [
+          {
+            prices: {
+              input_mtok: 2,
+              cache_write_mtok: 2.5,
+              cache_read_mtok: 0.2,
+              output_mtok: 10
+            }
+          },
+          {
+            constraint: {
+              start_date: "2026-09-01",
+              type: "start_date"
+            },
+            prices: {
+              input_mtok: 3,
+              cache_write_mtok: 3.75,
+              cache_read_mtok: 0.3,
+              output_mtok: 15
+            }
+          }
+        ]
+      },
       {
         id: "anubis-pro-105b-v1",
         name: "Anubis Pro 105B V1",
@@ -14816,7 +15120,14 @@ Compared to other leading proprietary and open-weights models Command A delivers
         id: "google/gemini-3.5-flash",
         name: "Gemini 3.5 Flash",
         match: {
-          equals: "google/gemini-3.5-flash"
+          or: [
+            {
+              equals: "google/gemini-3.5-flash"
+            },
+            {
+              regex: "^google/gemini-3\\.5-flash-\\d{8}$"
+            }
+          ]
         },
         prices: {
           input_mtok: 1.5,
@@ -19982,7 +20293,14 @@ Compared to other leading proprietary and open-weights models Command A delivers
         id: "x-ai/grok-4.3",
         name: "Grok 4.3",
         match: {
-          equals: "x-ai/grok-4.3"
+          or: [
+            {
+              equals: "x-ai/grok-4.3"
+            },
+            {
+              regex: "^x-ai/grok-4\\.3-\\d{8}$"
+            }
+          ]
         },
         prices: {
           input_mtok: 1.25,
@@ -22189,6 +22507,15 @@ Compared to other leading proprietary and open-weights models Command A delivers
             {
               equals: "grok-4.3"
             },
+            {
+              regex: "^grok-4\\.3-\\d{8}$"
+            },
+            {
+              equals: "x-ai/grok-4.3"
+            },
+            {
+              regex: "^x-ai/grok-4\\.3-\\d{8}$"
+            },
             {
               equals: "grok-4.3-latest"
             },
@@ -22564,7 +22891,7 @@ Compared to other leading proprietary and open-weights models Command A delivers
     ]
   }
 ];
-function Ye(i, o, u) {
+function Ze(i, o, u) {
   if (o <= 0) return 0;
   let s = i.base;
   for (const r of i.tiers)
@@ -22572,9 +22899,9 @@ function Ye(i, o, u) {
   return s * o / 1e6;
 }
 function C(i, o, u, s) {
-  return i === void 0 || o === void 0 ? 0 : typeof i == "number" ? i * o / 1e6 : Ye(i, o, s);
+  return i === void 0 || o === void 0 ? 0 : typeof i == "number" ? i * o / 1e6 : Ze(i, o, s);
 }
-function Ze(i, o) {
+function Ye(i, o) {
   let u = 0, s = 0;
   const r = i.input_tokens ?? 0, _ = i.cache_read_tokens ?? 0, h = i.cache_write_tokens ?? 0, k = i.cache_audio_read_tokens ?? 0, q = i.input_audio_tokens ?? 0, Q = i.output_audio_tokens ?? 0, B = o.cache_audio_read_mtok === void 0 ? 0 : k, G = o.cache_audio_read_mtok === void 0 && o.cache_read_mtok !== void 0 ? k : 0;
   let M = 0;
@@ -22620,12 +22947,12 @@ function Je(i, o) {
   }
   return i.prices[0].prices;
 }
-function D(i, o) {
-  return "or" in i ? i.or.some((u) => D(u, o)) : "and" in i ? i.and.every((u) => D(u, o)) : "equals" in i ? o.toLowerCase() === i.equals.toLowerCase() : "starts_with" in i ? o.toLowerCase().startsWith(i.starts_with.toLowerCase()) : "ends_with" in i ? o.toLowerCase().endsWith(i.ends_with.toLowerCase()) : "contains" in i ? o.toLowerCase().includes(i.contains.toLowerCase()) : "regex" in i ? new RegExp(i.regex).test(o) : !1;
+function F(i, o) {
+  return "or" in i ? i.or.some((u) => F(u, o)) : "and" in i ? i.and.every((u) => F(u, o)) : "equals" in i ? o.toLowerCase() === i.equals.toLowerCase() : "starts_with" in i ? o.toLowerCase().startsWith(i.starts_with.toLowerCase()) : "ends_with" in i ? o.toLowerCase().endsWith(i.ends_with.toLowerCase()) : "contains" in i ? o.toLowerCase().includes(i.contains.toLowerCase()) : "regex" in i ? new RegExp(i.regex).test(o) : !1;
 }
 function Xe(i, o) {
   const u = o.toLowerCase().trim(), s = i.find((r) => r.id === u);
-  return s || i.find((r) => r.provider_match && D(r.provider_match, u));
+  return s || i.find((r) => r.provider_match && F(r.provider_match, u));
 }
 function he(i, { modelId: o, providerApiUrl: u, providerId: s }) {
   if (s) {
@@ -22636,10 +22963,10 @@ function he(i, { modelId: o, providerApiUrl: u, providerId: s }) {
   if (u)
     return i.find((r) => new RegExp(r.api_pattern).test(u));
   if (o)
-    return i.find((r) => r.model_match && D(r.model_match, o));
+    return i.find((r) => r.model_match && F(r.model_match, o));
 }
 function et(i, o) {
-  return i.find((u) => D(u.match, o));
+  return i.find((u) => F(u.match, o));
 }
 function qe(i, o, u) {
   const s = et(i.models, o);
@@ -22665,7 +22992,7 @@ function tt(i, o, u) {
   if (!_) return null;
   const h = qe(_, s, ie);
   if (!h) return null;
-  const k = u?.timestamp ?? /* @__PURE__ */ new Date(), q = Je(h, k), Q = Ze(i, q);
+  const k = u?.timestamp ?? /* @__PURE__ */ new Date(), q = Je(h, k), Q = Ye(i, q);
   return {
     auto_update_timestamp: void 0,
     model: h,